Apurvakumar Pandya, أبورفاكومار بانديا،
Indian Institute of Public Health Gandhinagar (IIPHG), India مؤسسة الهند للصحة العامة غاندينagar (IIPHG)، الهند
REVIEWED BY تم مراجعتها من قبل
Daria Smirnova, داريا سميرنوفا،
Samara State Medical University, Russia جامعة سامارا الطبية، روسيا
Timur Syunyakov, تيمور سينياكوف،
Samara State Medical University, Russia جامعة سامارا الطبية، روسيا
*CORRESPONDENCE *الردود البريدية
Avinash De Sousa أفيناش دي سوسا
avinashdes888@gmail.com
received 25 August 2023 تلقيت في 25 أغسطس 2023
ACCEPTED 26 February 2024 مُفَعَّلَ 26 فبراير 2024
published 18 March 2024 نشرت في 18 مارس 2024
CITATION المُصْنِفَة
Thakkar A, Gupta A and De Sousa A (2024) ثاكار أ., غوبرت أ. و دي سوسا أ. (2024)
Artificial intelligence in positive mental health: a narrative review. الذكاء الاصطناعي في الصحة العقلية الإيجابية: مراجعة قصصية.
Front. Digit. Health 6:1280235. مجلة الصحة الرقمية الأمامية 6:1280235.
doi: 10.3389/fdgth.2024.1280235
Artificial intelligence in positive mental health: a narrative review الذكاء الاصطناعي في الصحة النفسية الإيجابية: مراجعة قصصية
Anoushka Thakkar ^(1){ }^{1}, Ankita Gupta ^(1){ }^{1} and Avinash De Sousa ^(2){ }^{2} أنوشكا ثاككار ^(1){ }^{1} ، أنكيتا غوتا ^(1){ }^{1} وأفيناش دي سوسا ^(2){ }^{2}^(1){ }^{1} MINDLab, The MINDS Foundation, Mumbai, India, ^(2){ }^{2} Desousa Foundation, Mumbai, India ^(1){ }^{1} مينكلاب، مؤسسة مينكلز، مومباي، الهند، ^(2){ }^{2} مؤسسة دي سوسا، مومباي، الهند
Abstract مقدمة
The paper reviews the entire spectrum of Artificial Intelligence (AI) in mental health and its positive role in mental health. Al has a huge number of promises to offer mental health care and this paper looks at multiple facets of the same. The paper first defines AI and its scope in the area of mental health. It then looks at various facets of AI like machine learning, supervised machine learning and unsupervised machine learning and other facets of AI. The role of AI in various psychiatric disorders like neurodegenerative disorders, intellectual disability and seizures are discussed along with the role of AI in awareness, diagnosis and intervention in mental health disorders. The role of AI in positive emotional regulation and its impact in schizophrenia, autism spectrum disorders and mood disorders is also highlighted. The article also discusses the limitations of AI based approaches and the need for AI based approaches in mental health to be culturally aware, with structured flexible algorithms and an awareness of biases that can arise in AI. The ethical issues that may arise with the use of AI in mental health are also visited. المراجعة تقييم نطاق كامل من الذكاء الاصطناعي (AI) في الصحة العقلية ودوره الإيجابي في الصحة العقلية. يقدم الذكاء الاصطناعي عددًا كبيرًا من الوعود لتقديم الرعاية الصحية العقلية وهذا المقال يتعمق في جوانب متعددة من نفس الشيء. يبدأ المقال بتعريف الذكاء الاصطناعي ودوره في مجال الصحة العقلية. ثم يتعمق في جوانب مختلفة من الذكاء الاصطناعي مثل التعلم الآلي، التعلم الآلي الموجه والتعلم الآلي غير الموجه وأخرى من جوانب الذكاء الاصطناعي. يتم مناقشة دور الذكاء الاصطناعي في أمراض نفسية مختلفة مثل الأمراض العصبية التدريجية، الإعاقة الذهنية والصرع، بالإضافة إلى دور الذكاء الاصطناعي في الوعي وال تشخيص وتدخل الأمراض النفسية. يتم التأكيد على دور الذكاء الاصطناعي في تنظيم العواطف الإيجابية وتأثيره في الفصام، اضطرابات طيف التوحد والاضطرابات المزاجية. يتناول المقال أيضًا القيود للطرق القائمة على الذكاء الاصطناعي والضرورة لطرق الذكاء الاصطناعي في الصحة العقلية أن تكون حساسة للثقافة، مع خوارزميات مرنة منظمة وعي بتحيزات التي يمكن أن تنشأ في الذكاء الاصطناعي. يتم أيضًا مناقشة القضايا الأخلاقية التي قد تنشأ مع استخدام الذكاء الاصطناعي في الصحة العقلية.
KEYWORDS كلمات مفتاحية
artificial intelligence, positive mental health, ethics, machine learning, mental health الذكاء الاصطناعي، الصحة العقلية الإيجابية، الأخلاقيات، التعلم الآلي، الصحة العقلية
Introduction مقدمة
“What is AI?” This has been a popular question since the inception of AI or Artificial Intelligence as a discipline in the 1950s (1). There are many proposed definitions of AI, however, most of them are aligned around the concept of machines capable of human behaviors or creating computer programs (2). The founding father of this discipline, John McCarthy (3) described this process as “that of making a machine behave in ways that would be called intelligent if a human were so behaving.” The goal of AI, according to him, was to develop machines that behave as though they were intelligent. "ما هو الذكاء الاصطناعي؟" هذا السؤال كان شائعاً منذ نشأة الذكاء الاصطناعي أو الذكاء الاصطناعي كعلم في الخمسينيات (1). هناك العديد من التعريفات المقترحة للذكاء الاصطناعي، ولكن معظمها تتوافق حول المفهوم من الآلات القادرة على التصرف بشكل بشري أو إنشاء برامج حاسوبية (2). كان مؤسس هذا العلم، جون مارشال (3)، يصف هذه العملية بأنها "عملية جعل الآلة تصرف نفسها بطريقة سيُسمى بها ذكاءً إذا كان الإنسان يتصرف هكذا." وقال بأن هدف الذكاء الاصطناعي، هو تطوير آلات تصرف كأنها ذكية.
Over the last couple decades, AI has evolved-from playing with toy problems like chess, to acquiring new skills and seeking to discover its own limits. After 60 years, AI has made its way into industries and the consciousness of people (4). It is only now in the 21st century that this discipline has transformed every aspect of our life in such a significant way, that it is referred to as the “Age of AI” (5). على مدى العقود الأخيرة، تطور الذكاء الاصطناعي - من اللعب بمشاكل صغيرة مثل الشطرنج، إلى الحصول على مهارات جديدة و محاولة اكتشاف حدود نفسه. بعد 60 عاماً، وصل الذكاء الاصطناعي إلى الصناعات وعلى وعي الناس (4). ولم يكن حتى الآن في القرن الحادي والعشرين أن تحول هذا العلم كل جانب من جوانب حياتنا بدرجة كبيرة، لدرجة أنه يُعرف بـ "العصر الذكي" (5).
Unlike computer science, AI has evolved with a more fluid definition, owing to the different conceptions of intelligence. It was conceived as a computer system that is similar to the human mind in numerous ways (6). AI is an umbrella term that encompasses a wide range of approaches and techniques to develop computational systems which perform cognitive processes and tasks that are characteristic to humans. Examples of such processes and tasks include learning, reasoning, problem solving, drawing inferences and generalization (2). بما عدا علم الحاسوب، تطورت الذكاء الاصطناعي بتعريف أكثر مرونة، وذلك بسبب المفاهيم المختلفة للذكاء. لقد تم conceivingه كنظام حاسوب يشبه الدماغ البشري في العديد من الطرق (6). الذكاء الاصطناعي هو مصطلح شامل يغطي مجموعة واسعة من المناهج والخوارزميات لتطوير أنظمة حاسوبية تقوم بإجراء عمليات وعمليات تتوافق مع الإنسان. أمثلة لعمليات وعمليات كهذه تشمل التعلم، المنطق، حل المشكلات، استنتاج والعمليات العامة (2).
Background, history and evolution of Al الخلفية، التاريخ والتطور للذكاء الاصطناعي
The history of AI dates back to the 1950s, a time marked by the initial development of machines with the acumen for human-like decision making and reasoning. This period witnessed significant landmarks, notably the introduction of Unimate in 1961, an industrial robot arm, as well as the creation of Eliza in 1964, a communication-based chatterbot. These headways worked towards laying the bedrock for future strides in robotics and AI. However, in spite of progress in the engineering field during this time, the medical field was slow to adopt AI (7). تاريخ الذكاء الاصطناعي يعود إلى عقد الخمسينيات، وهو وقت تميز بالتطوير الأولي للآلات بمهارة في اتخاذ قرارات ومنطق يشبه البشر. شهد هذا العصر نقاط تحولية مهمة، بما في ذلك إطلاق Unimate في عام 1961، وهو رأس حاسوب صناعي، بالإضافة إلى إنشاء Eliza في عام 1964، وهو روبوت محادثي يعتمد على التواصل. هذه التقدمات ساهمت في وضع الأساس لخطوات مستقبلية في الروبوتات والذكاء الاصطناعي. ولكن، على الرغم من التقدم في مجال الهندسة خلال هذا الوقت، كان مجال الطب بطيئًا في تبني الذكاء الاصطناعي (7).
Modern concept of AI emerged in the mid-20th century, when Alan Turing, a scholar in this field laid down the theoretical groundwork for AI. This was achieved through his pioneering work on the “Turing Test” and his concepts of universal computing machines (8). The term AI was first coined by John McCarthy in 1956 as “the science and engineering of making intelligent machines,” marking the establishment of the field as a distinct discipline (9). فكرة الذكاء الاصطناعي الحديثة ظهرت في منتصف القرن العشرين، عندما وضع عالم في هذا المجال، ألان تورنج، أسس теорيتية للذكاء الاصطناعي. تم تحقيق ذلك من خلال عمله الرائد على "اختبار تورنج" ومفاهيمه حول الأجهزة الحاسوبية العامة (8). تم ابتكار مصطلح الذكاء الاصطناعي لأول مرة من قبل جون مارشال في عام 1956 كـ "العلوم والهندسة في صناعة الآلات الذكية"، مما علام على تأسيس المجال كدعامة منفصلة (9).
The progress of AI can be characterized by a number of critical milestones. The Darthmouth Workshop in 1956, which is often considered to be the birthplace of AI, brought together likeminded researchers to explore the possibilities of developing machines mimicking human intelligence. The next couple of decades witnessed some more breakthroughs in this area of research, with systems to manipulate symbols and follow logical rules being designed (10). يمكن وصف تقدم الذكاء الاصطناعي ببعض المحطات الحرجة. ورشة عمل دارتموث في عام 1956، التي يُعتبر غالباً مكان ميلاد الذكاء الاصطناعي، جمعت معاً الباحثين ذوي التفاهمة المشتركة لاستكشاف إمكانيات تطوير أجهزة تقلد الذكاء البشري. شهد العقود التالية بعض الاكتشافات المهمة في هذا المجال من البحث، مع تصميم أنظمة لمعالجة الرموز والالتزام بالقواعد المنطقية (10).
In the 1980s, we saw a rise in expert systems, which aimed at capturing human expertise in specific domains using rule-based systems (11). Limitations in dealing with the uncertainty and complexity of real-world situations however, lead to the AI winter. This was a period of dampened enthusiasm and reduced funding in the field of AI which lasted from the 1980s to early 1990s (12). في الثمانينيات، شهدنا ارتفاعاً في أنظمة الخبراء، والتي أرادت التمسك بخبرة البشر في مجالات محددة باستخدام أنظمة قائمة على القواعد (11). ومع ذلك، القيود في التعامل مع عدم اليقين ومعقدة المواقف الحقيقية أدت إلى فصل الشتاء في الذكاء الاصطناعي. كان هذا فترة من تلاشي الحماس وتقليل التمويل في مجال الذكاء الاصطناعي، والتي استمرت من الثمانينيات إلى أوائل التسعينيات (12).
The resurgence of AI in the late 20th century can be attributed to the advancements in machine learning, like the ability of neural networks to model complex patterns. With the field of machine learning gaining momentum and the development of algorithms, the concept of neural networks showed a rise in the 1990s (13). الانتعاش الذكاء الاصطناعي في أواخر القرن العشرين يمكن تفسيره بفضل التقدم في التعلم الآلي، مثل القدرة على شبكات العصبونية على بناء الأنماط المعقدة. مع تطور مجال التعلم الآلي وتطوير الخوارزميات، أظهر مفهوم الشبكات العصبونية ارتفاعاً في التسعينيات (13).
The 21st century witnessed a crucial breakthrough in AI as a result of the availability of improved computational resources and massive datasets. Deep learning techniques, like language processing and image recognition revolutionized fields like language understanding and computer vision (14). Seminal advancements in AI including IBM’s Watson, an open-domain question-answering system that won a game show in 2011, allowed for the rapid growth of AI in medicine as well as other industries. شهد القرن الحادي والعشرين انطلاقة حاسمة للذكاء الاصطناعي كنتيجة لوجود موارد حسابية مُحسّنة ومجموعات بيانات ضخمة. تقنيات التعلم العميق، مثل معالجة اللغات والتعرف على الصور، غيرت مجالات مثل فهم اللغة والرؤية الحاسوبية (14). التقدم الكبير في الذكاء الاصطناعي، بما في ذلك واتسون من شركة آي بي إم، نظام إجابة الأسئلة المتعددة المجالات الذي فاز في برنامج تلفزيوني في عام 2011، سمح بتوسع سريع للذكاء الاصطناعي في الطب والصناعات الأخرى.
Use of AI in industries استخدام الذكاء الاصطناعي في الصناعات
AI has grown exponentially as a transformative technology and is increasingly present in our daily lives-across different industries (15). AI is changing traditional processes, enhancing decision making capabilities and unlocking newer avenues for novelty and innovation. As AI continues to evolve and reshape the way لقد نمى الذكاء الاصطناعي بسرعة كبيرة كтехнологية تحويلية ويمثل حضوراً متزايداً في حياتنا اليومية عبر مختلف الصناعات (15). يغير الذكاء الاصطناعي العمليات التقليدية، ويعزز قدرات اتخاذ القرار، ويفتح آفاقاً جديدة للإبتكار والابتكار. مع استمرار تطور الذكاء الاصطناعي وتغيير الطريقة
business is conducted, wecan expect a larger shift from industries to a landscape with intelligent systems that are driven by machine learning techniques and advanced algorithms. عندما يتم إدارة الأعمال، نتوقع تحولاً أكبر من الصناعات إلى المشهد المكون من أنظمة ذكية تُدار باستخدام تقنيات التعلم الآلي والخوارزميات المتقدمة.
The integration of AI into industries consists of a wide spectrum, from healthcare to recruitment, logistics, education, transportation security, e-government, and public sectors. Current advances in the development and applications of AI systems, such as machine learning, deep learning, autonomous or semi-autonomous systems, reshape science and society, creating new opportunities in the way we live, work, travel and do business. تتضمن دمج الذكاء الاصطناعي في الصناعات نطاقاً واسعاً، من الرعاية الصحية إلى التوظيف، واللوجستيات، والتعليم، والأمان في النقل، والحكومة الإلكترونية، والقطاعات العامة. التطورات الحالية في تطوير وتطبيق أنظمة الذكاء الاصطناعي، مثل التعلم الآلي، والتعلم العميق، وأنظمة ذاتية أو نصف ذاتية، تُعيد تشكيل العلم وال xã hội، مما يخلق فرصاً جديدة في الطريقة التي نعيش، ونعمل، وننتقل، ونقوم بالإدارة.
Financial institutions have been employing AI algorithms to detect any fraudulent activities, risk assessment and mitigation, and for enhancing security and stability within the sector (16). In healthcare, AI-powered diagnostic tools are leveraging image recognition and natural language processing to assist medical professionals in accurate disease detection and treatment planning (17). The manufacturing industry has witnessed the implementation of AI-driven maintenance, which in turn has led to a reduction in the downtime and optimizing production processes (18). Transportation and logistics are benefiting from AI-powered route optimization and autonomous vehicles which promise a safer and more efficient mobility solution (19). البنوك المالية تستخدم خوارزميات الذكاء الاصطناعي لتحديد أي أنشطة احتيالية، وتقييم المخاطر وتخفيفها، وتعزيز الأمان والاستقرار داخل القطاع (16). في الرعاية الصحية، أدوات التشخيص المدعومة بالذكاء الاصطناعي تستفيد من التعرف على الصور ومعالجة اللغة الطبيعية لمساعدة المختصين الطبيين في اكتشاف دقيق للأمراض وتخطيط العلاج (17). شهد قطاع التصنيع تطبيقاً للصيانة المدعومة بالذكاء الاصطناعي، مما أدى إلى تقليل وقت التوقف وتحسين عمليات الإنتاج (18). يفيد قطاع النقل واللوجستيات من تحسين المسارات المدعوم بالذكاء الاصطناعي والسيارات الذاتية القيادة التي تعد حلولاً أكثر أماناً وفعالية للنقل (19).
The introduction of AI into industries also empowers decisionmakers with a more informed decision based on data-driven insights. This in turn enables more informed and strategic choices. Machine learning algorithms for example can process vast volumes of data to uncover patterns and trends, facilitating predictive and prescriptive analytics that drive operational efficiency (20). Through its functioning, AI contributes to resource optimization and cost reduction, ultimately bolstering competitiveness. إدخال الذكاء الاصطناعي في الصناعات يعزز أيضاً قرارات المقررين بقرارات أكثر تحديداً استناداً إلى فهمات مدعومة بالبيانات. وهذا بدوره يتيح خيارات أكثر تحليلاً واستراتيجية. على سبيل المثال، يمكن خوارزميات التعلم الآلي معالجة كميات هائلة من البيانات لاستكشاف الأنماط والاتجاهات، مما يسهل التحليل التنبؤي والتحليل التوصيفي الذي يعزز كفاءة العمليات (20). من خلال أدائها، يساهم الذكاء الاصطناعي في تحسين الموارد وتقليل التكاليف، مما يعزز في نهاية المطاف المنافسة.
A popular form of AI includes machine learning, which learns from imputed data or from the machines experiences and is then taught values via conditioning (21). This can be used in various capacities like voice recognition software (example Alexa and Siri), to medical diagnosis through pattern recognition techniques (22). Other examples of AI in various industries includes AI rapper FN Meka who has signed to Capitol Records and research at Columbia University that resulted in Teaching robots to visual themselves through AI (23). In a 2022 survey consisting of 850 organizations and 18 geographies, 77%77 \% reported prioritizing AI regulations as company-wide policies and 80%80 \% stated they would invest in the development of ethical AI (24). صيغة شائعة من الذكاء الاصطناعي تشمل التعلم الآلي، الذي يتعلم من البيانات المُدخلة أو من تجارب الآلات ثم يُعلم القيم عبر التشكيل (21). يمكن استخدامها في أدوار متعددة مثل برمجيات التعرف على الصوت (مثال أليكسا وسيري)، ل تشخيص الأمراض الطبية من خلال تقنيات التعرف على الأنماط (22). أمثلة أخرى من الذكاء الاصطناعي في الصناعات المختلفة تشمل الرapper الذكاء الاصطناعي FN Meka الذي подписى مع Capitol Records وبحث في جامعة كولومبيا الذي أدى إلى تعليم الروبوتات كيفية رؤية أنفسهم من خلال الذكاء الاصطناعي (23). في استطلاع عام 2022 شمل 850 منظمة و 18 منطقة، أشارت 77%77 \% إلى أنهم يُعليون من أولويات تنظيم الذكاء الاصطناعي كسياسات شاملة للشركة، وأشارت 80%80 \% إلى أنهم سينفقون في تطوير الذكاء الاصطناعي الأخلاقي (24).
Just like most other fields, mental health care too has been impacted by the revolution in digital technology, specifically AI. The emergence of digital mental health as a sub-field signifies the integration of technology into mental health care practices, and within this domain, AI-driven solutions have become an important contributor to positive mental health outcomes (25). The mental health landscape has witnessed a paradigm shift over the last couple years, with technological advancements playing an important role in enhancing accessibility, personalization, and efficacy of interventions. The application of AI in positive mental health is multifaceted, ranging from the development of applications and websites that make use of chatbots and virtual assistants to offer immediate mental health support, someone to مثل معظم المجالات الأخرى، تأثرت رعاية الصحة العقلية أيضاً بالثورة في التكنولوجيا الرقمية، خاصة بالذكاء الاصطناعي. ظهور الصحة العقلية الرقمية ك مجال فرعي يشير إلى دمج التكنولوجيا في ممارسات رعاية الصحة العقلية، وفي هذا المجال، أصبحت الحلول المبنية على الذكاء الاصطناعي مساهماً مهماً في نتائج الصحة العقلية الإيجابية (25). شهد المشهد الصحي العقلاني تحولاً جذرياً على مدى السنوات القليلة الأخيرة، مع دور التقدم التكنولوجي الهام في تحسين الوصول، والخصوصية، والفعالية للإجراءات. تطبيق الذكاء الاصطناعي في الصحة العقلية الإيجابية متعدد الأوجه، يتراوح من تطوير تطبيقات وأماكن إلكترونية تستخدم روبوتات محادثة ومساعدات افتراضية لتقديم الدعم الصحي العقلاني فوريًا، شخصًا لـ
talk to in times of distress and identifying potential mental health issues. Machine learning algorithms are also being used to tailor personalized therapeutic interventions, while monitoring realtime data through wearable devices for better informed decisionmaking. In the context of positive mental health, AI-driven interventions extend beyond clinical settings to wellness applications, delivering tailored recommendations and mindfulness exercises. The purpose of this review is to focus on the ways AI is being used in positive mental health. تحدث إلى في أوقات الأزمة واكتشاف المشكلات العقلية المحتملة. كما يتم استخدام خوارزميات التعلم الآلي لتخصيص تدخلات العلاج الشخصية، بينما يتم مراقبة بيانات thời gian واقعي عبر أجهزة اللباس لاتخاذ قرارات أكثر تحديداً. في سياق الصحة العقلية الإيجابية، تدخل التدخلات الموجهة بالذكاء الاصطناعي ما وراء البيئات الطبية إلى تطبيقات الصحة والنمو، حيث توفر توصيات مخصصة وممارسات الانتباه الذهني. الغرض من هذا المراجعة هو التركيز على الطرق التي يتم استخدام الذكاء الاصطناعي فيها في الصحة العقلية الإيجابية.
Understanding artificial intelligence and its application in mental health care فهم الذكاء الاصطناعي وتطبيقه في رعاية الصحة العقلية
Artificial intelligence and its key components الذكاء الاصطناعي وأهم مكوناته
Artificial Intelligence (AI) refers to the field of computer science and technology that aims to create systems and machines capable of performing tasks that typically require human intelligence, such as understanding natural language, recognizing patterns, making decisions, learning from experience, and adapting to new situations, all achieved through a combination of algorithms, data processing, and iterative improvement processes (9). Following are the notable key components of AI. يُعرف الذكاء الاصطناعي (AI) بأنه مجال من علوم الكمبيوتر والتكنولوجيا يهدف إلى إنشاء أنظمة ومعدات قادرة على أداء مهام عادة ما تتطلب ذكاءً بشريًا، مثل فهم اللغة الطبيعية، التعرف على الأنماط، اتخاذ القرارات، التعلم من التجربة، والتكيف مع مواقف جديدة، كل ذلك يتم تحقيقه من خلال مزيج من الخوارزميات، معالجة البيانات، وعمليات التحسين المتكررة (9). تلي أدناه المكونات الرئيسية المهمة للذكاء الاصطناعي.
Machine Learning (ML) is a subset of AI that involves the development of algorithms that allow systems to learn from data and improve their performance over time without being explicitly programmed (26). It encompasses various techniques like supervised learning, unsupervised learning, deep learning, and reinforcement learning. ML algorithms can analyze patient data, such as electronic health records and behavioral patterns, to assist in diagnosing mental health conditions like depression, anxiety, and schizophrenia (26). They can also predict the risk of developing certain disorders based on historical data. التعلم الآلي (ML) هو جزء من الذكاء الاصطناعي يشمل تطوير خوارزميات تسمح للأنظمة بتعلم من البيانات وتحسين أدائها على مدى الوقت دون برمجة صريحة (26). إنه يشمل تقنيات متنوعة مثل التعلم الموجه، التعلم غير الموجه، التعلم العميق، والتعلم المعزز. يمكن للخوارزميات التعليمية تحليل بيانات المرضى، مثل سجلات الرعاية الصحية الإلكترونية والأنماط السلوكية، لمساعدة تشخيص الحالات الصحية العقلية مثل الاكتئاب، القلق، والذهان (26). يمكنها أيضاً توقع مخاطر تطور أمراض معينة بناءً على البيانات التاريخية.
Additionally, they can help tailor treatment plans by analyzing patient data to recommend specific therapies, medications, or interventions based on individual characteristics and response patterns (25). علاوة على ذلك، يمكنها المساعدة في تخصيص خطط العلاج عن طريق تحليل بيانات المرضى لتوصية بالعلاجات المحددة، أو الأدوية، أو التدخلات بناءً على الخصائص الفردية والأنماط الاستجابية (25).
In Supervised Machine Learning (SML), the data are prelabeled beforehand, for example, distinguishing between the diagnosis of major depressive disorder (MDD) and the absence of depression. The algorithm acquires the capacity to establish the most accurate between input features extracted from various sources of data, including sociodemographic, biological, and clinical measures (27). This intricate linkage of features facilitates the algorithm’s adeptness in generating highly accurate predictions within these designated categories (28). It is important to note that the labels can be categorical (i.e., MDD or not or continuous (i.e., along the spectrum of severity) in nature. The machine undergoes supervised ML because the labeled data serves as guidance (“teacher”) for the algorithm, teaching it how to assign labels to the data, enabling the algorithm to learn how to link features with specific labels; subsequently, having learned from extensive labeled training data, the algorithm is assessed في التعلم الآلي الموجه (SML)، يتم تسجيل البيانات مسبقًا، على سبيل المثال، من خلال تمييز بين تشخيص الاضطراب الاكتئابي المزمن (MDD) وغياب الاكتئاب. يكتسب الخوارزمية القدرة على تحديد الأكثر دقة بين الميزات المشتقة من مصادر بيانات مختلفة، بما في ذلك المتعلق بالديموغرافيا الاجتماعية، والبيولوجية، والقياسات الطبية (27). هذا الروابط المعقدة للميزات يسهل مهارة الخوارزمية في إنشاء توقعات دقيقة جدًا ضمن هذه الفئات المحددة (28). من المهم ملاحظة أن العلامات يمكن أن تكون فئوية (أي، MDD أو ليس كذلك) أو متصلة (أي، على طول طيف شدة) في الطبيعة. يعتمد الآلة على التعلم الآلي الموجه لأن البيانات المُسجلة تعمل كدليل ("معلم") للخوارزمية، تدريبه على كيفية تخصيص العلامات للبيانات، مما يتيح للخوارزمية تعلم كيفية ربط الميزات بعلامات محددة؛ وبعد تعلمها من بيانات التدريب المُسجلة على نطاق واسع، يتم تقييم الخوارزمية
using unlabeled test data to ascertain its ability to accurately classify the desired outcome, such as MDD (25). باستخدام بيانات الاختبار غير المُسجلة لتحديد قدرتها على تصنيف النتيجة المطلوبة بدقة، مثل MDD (25).
Conversely, in Unsupervised Machine Learning (UML), the algorithms function without predefined labels, enabling them to detect similarities among input features and unveil the inherent data patterns. However, they lack the ability to associate features with a known label (29). The absence of labels presents a greater challenge for UML; however, it offers the advantage of uncovering the inherent structure within a dataset without significant pre existing biases. For instance, extensive feature datasets, such as neuroimaging biomarkers, possess the potential to unveil insights about unknown subtypes within psychiatric disorders like schizophrenia. UML can aid in identifying groups of biomarkers that define these variations, thereby contributing to prognosis and optimal treatment strategies (25). في المقابل، في التعلم الآلي غير الموجه (UML)، تعمل الخوارزميات بدون علامات مسبقة، مما يتيح لها اكتشاف التشابهات بين الميزات المُدخلة وتكشف الأنماط الخفية للبيانات. ولكنها تفتقر القدرة على ربط الميزات بعلامة معروفة (29). غياب العلامات يمثل تحديًا أكبر لـ UML؛ ومع ذلك، يوفر مزيدًا من المزايا من خلال كشف الهيكل الخفي ضمن مجموعة البيانات دون تحيزات مسبقة كبيرة. على سبيل المثال، مجموعات البيانات الميزية الواسعة، مثل المؤشرات الحيوية للتصوير العصبي، تمتلك القدرة على كشف أسرار حول الأنواع الفرعية غير المعروفة للاضطرابات العقلية مثل الفصام. يمكن لـ UML المساعدة في تحديد مجموعات المؤشرات الحيوية التي تحدد هذه التغيرات، مما يساهم في التنبؤ بالنتائج وتطوير استراتيجيات العلاج الأمثل (25).
Artificial Neural Networks (ANNs) and Deep Learning (DL) are based on the structure and functioning of the human brain, DL involves handling complex and raw data, without human guidance, by employing ANNs that resemble the manner in which a human brain thinks (30). This enables them to model complex patterns and relationships in data processed through multiple “hidden” layers (30-32). Deep learning techniques can be used for image analysis in brain scans (MRI, CT, etc.), aiding in the identification of structural abnormalities linked to mental health disorders. الشبكات العصبونية الاصطناعية (ANNs) والتعلم العميق (DL) مبنية على هيكل ووظيفة الدماغ البشري، حيث يشمل التعلم العميق التعامل مع بيانات معقدة ومتدخلة بدون إشراف بشري، وذلك باستخدام شبكات عصبونية اصطناعية تشبه طريقة تفكير الدماغ البشري (30). وهذا يتيح لهما تصور الأنماط والروابط المعقدة في البيانات المعالجة عبر طبقات "خفية" متعددة (30-32). يمكن استخدام تقنيات التعلم العميق لتحليل الصور في مسح الدماغ (مثل MRI و CT وغيرها)، مما يساعد في تحديد العيوب الهيكلية المرتبطة باضطرابات الصحة العقلية.
Natural Language Processing (NLP) is another subfield of AI that enables machines to process, comprehend, interpret, and generate human language. Researchers refine and make use of such tools in real-world applications, creating spoken dialogue systems and speech-to-speech translation engines, mining social media for information about health or finance, and identifying sentiment and emotion toward products and services (33). NLP can significantly aid in analyzing written or spoken language to detect emotional states and changes, assisting clinicians monitor patients’ mental well-being through texts, chats, or speech. Furthermore, the integration of AI-powered chatbots amplifies this capacity by engaging users in text-based dialogues, thereby delivering real-time assistance, coping mechanisms, and even recommendations for consultation with mental health professionals. التعامل مع اللغة الطبيعية (NLP) هو مجال فرعي آخر من الذكاء الاصطناعي الذي يتيح للآلات معالجة وفهم وتفسير وتوليد اللغة البشرية. يصل الباحثون إلى تحسين واستخدام هذه الأدوات في التطبيقات العملية، حيث يخلقون أنظمة диалогية شفهية ومحركات ترجمة صوت إلى صوت، ويحللون وسائل التواصل الاجتماعي بحثاً عن معلومات صحية أو مالية، ويتتبعون المزاج والعواطف تجاه المنتجات والخدمات (33). يمكن للـ NLP المساعدة بشكل كبير في تحليل اللغة المكتوبة أو الشفهية لتحديد الحالات العاطفية والتغيرات، حيث يساعد الأطباء المعالجين على مراقبة الصحة العقلية للمرضى من خلال النصوص أو الحوارات أو الكلام. علاوة على ذلك، دمج روبوتات الحوار المدعومة بالذكاء الاصطناعي يزيد من هذه القدرة عن طريق مشاركة المستخدمين في الحوارات النصية، مما يوفر مساعدة فورية، وسبل التعامل مع المواقف، وحتى توصيات للإشارة إلى استشارة المختصين في الصحة العقلية.
Reinforcement Learning (RL) is a type of ML algorithm that involves training an agent to make sequential decisions by interacting with an environment. The algorithm acts as an agent in an interactive environment that learns by trial and error using rewards from its own actions and experiences (25). RL offers a dynamic framework for tailoring therapeutic interventions. For instance, AI-driven applications, such as virtual reality exposure therapy, can be designed to adjust exposure levels in response to patient reactions. This ensures that the intensity of exposure is optimized to minimize distress while still facilitating progress (34). The AI agent learns from patient feedback, continually refining the intervention strategy over time. التعلم المعزز (RL) هو نوع من خوارزميات التعلم الآلي التي تتضمن تدريب عميل لاتخاذ قرارات متتالية من خلال التفاعل مع بيئة. تقوم الخوارزمية بعمل عميل في بيئة تفاعلية تعلم بالمحاولة والخطأ باستخدام المكافآت من أفعاله وخبراته (25). يوفر التعلم المعزز إطار ديناميكي لتخصيص تدخلات العلاجية. على سبيل المثال، التطبيقات الموجهة بالذكاء الاصطناعي، مثل العلاج بالانغماس في الواقع الافتراضي، يمكن تصميمها لتعديل مستويات الانغماس استجابةً لاستجابات المريض. وهذا يضمن أن قوة الانغماس تكون مُحسّنة لتقليل الشعور بالضيق مع الحفاظ على تسهيل التقدم (34). يتعلم العميل الذكاء الاصطناعي من ملاحظات المريض، وتدعيم الاستراتيجية العلاجية بشكل مستمر على مر الوقت.
Computer Vision is an AI subfield that encompasses the development of computational systems that facilitate the interpretation and comprehension of visual data from the environment, similar to how humans perceive images and videos. Its practical implementations include image recognition, object رؤية الحاسوب هو مجال فرعي من الذكاء الاصطناعي يشمل تطوير نظم حاسوبية تسهل تفسير وفهم البيانات البصرية من البيئة، تمامًا مثل الطريقة التي يعبر بها البشر عن الصور والفيديوهات. تطبيقاته العملية تشمل التعرف على الصور والكائنات