Applications of AI in Medical Diagnostics

Machine vision is emerging in medical diagnostic applications, and it should be noted that improvements in that field will correlate to the successful implementation of AI and ML in healthcare applications. However, the process of trial and error will significantly influence the value of this technology in the real world. The emerging Machine Learning based diagnostics tools mainly fall under four major categories:

-Oncology : Researchers are using deep learning to train algorithms to recognize cancerous tissues to aid trained physicians.

-Pathology : Pathology is the medical specialty that is concerned with the diagnosis of disease based on the laboratory analysis of bodily fluids and tissues. Machine learning technologies can enhance the efforts of pathologists with microscopes.

-Chatbots : AI-based chatbots with speech recognition capability can identify patterns in symptoms of a patient to form a potential diagnosis, preventing disease and/or recommend an appropriate course of action.

-Rare diseases : Facial recognition software using ML helps physicians to diagnose rare diseases. Faces in photos of patients are analyzed and with deep learning algorithms, it can detect phenotypes that correlate with rare genetic diseases.

Oncology

Stanford University researchers have trained an algorithm to diagnose skin cancer using deep learning, specifically deep convolutional neural networks (CNNs). The algorithm was trained to detect skin cancer or melanoma using “130,000 images of skin lesions representing over 2,000 different diseases.”

The US have approximately 5.4 million new skin cancer diagnoses each year and the early detection is critical for a greater rate of survival. Early detection correlates with a 97% five-year survival rate but quickly decreases with later stages, hitting the 15–20 percent margin at stage IV. A visual examination is usually the first step of a skin cancer diagnosis and a dermatologist inspects a lesion of interest and if the initial evaluation is inconclusive, the dermatologist will follow up with a biopsy.

Stanford’s deep learning algorithm was tested against 21 board-certified dermatologists, results showed that the algorithm had the same ability as the 21 dermatologists in determining the best course of action in analyzing the sample images. Even though the results are promising, the research team acknowledges that additional, rigorous testing is required before the algorithm can be integrated into clinical practice.

Perfomatix helped develop a cancer diagnostics that use AI for image processing and have an intuitive dashboard that provides annotations on high-resolution histopathology images, patient history as well as decision support systems for diagnosing cancer. Read all about it here.

Chatbots

Speech recognition in chatbots can compare to the symptoms that it receives from a user against a database of diseases. Chatbots using AI could recommend an appropriate course of action by analyzing the symptoms, patient history, and patient circumstances.

In sympomatic analysis by chatbots, for a simple flu-like symptoms chatbots might a recommend to visit the pharmacy for over-the-counter medication. For more serious symptoms, the bot could recommend dialing an emergency number or directly visiting a healthcare facility.

In some ways, chatbots in the healthcare setting are a bit like IoT for the fitness space a few years ago. It seems inevitable, it seems to have great promise. In the 2020s, we surely predict chatbot based healthcare apps will have many successful applications of AI in healthcare.

Pathology

Pathologists diagnose diseases by manually observing images under a microscope, this technique has remained relatively unchanged for over a century. This often lacks efficiency, AI and ML powered pathological diagnosis can offer improved speed and accuracy. A research by Harvard Medical School have used deep learning to train an algorithm capable of integrating multiple speech recognition and image recognition to diagnose tumors.

The researchers began with hundreds of images with labeled regions showing cancerous and noncancerous cells. The labeled regions were then extracted resulting in millions of examples that served as the basis for the model that would train the algorithm.

When compared to human pathologists, study results showed that the algorithm achieved a diagnostic success rate of 92 percent; four percentage points lower than the human rate of 96 percent. However, when the algorithm and human results are combined, an accuracy rate of 99.5 percent was achieved.

This shines light on the much needed speed that’s required in pathogenic diagnosis. AI and ML can help pathologists to achieve efficiency in a global scale.

Rare Diseases

A facial recognition software Face2Gene is an exemplary solution in this category that uses machine learning to help clinicians diagnose rare diseases (in this case, genetically associated facial dysmorphic features). Facial analysis and deep learning are used to detect phenotypes that correlate with rare genetic diseases. The platform is currently available only to US based trained clinicians to prevent false positives and supports over 7,500 disorders.

Summing up

Applications of AI and ML based solutions in medical diagnostics are in the early adoption phase across the globe. Some that are available have limited access, hence the data on outcomes are not enough to draw an informed conclusion regarding the efficiency. But, we can widely acknowledge the potential of AI and ML in healthcare diagnosis to impact clinics, practicioners and health care systems. It has the ability to revolutionize the industry for individuals to understand their health in real-time.

Continued rigorous testing of these applications will be necessary to validate their utility combined with education of clinicians and healthcare systems around how to effectively implement these technologies in clinical practice. The huge investments in healthcare AI sector suggest the next wave of medical diagnostic tech is fast approaching. Healthcare solutions companies make great efforts to bring accurate and reliable medical diagnostics based on machine and deep learning applications to market.

That being said, AI in medical diagnostics are taken with great caution among the healthcare workers. Many clinicians are largely unsatisfied and still questions the reliability and sensitivity of machines.

The practical integration of such smart machines should be implemented without undermining human expertise and clinical care.