One Student Found Purpose Beyond Traditional Medicine
Sungjoon Hong entered medical school with a clear desire to help vulnerable patients recover. He found deep personal fulfillment through support during difficult moments in people’s lives. Pain management and physical medicine and rehabilitation matched that purpose especially well. Those interests eventually led him toward broader questions about patient recovery and care.
Artificial intelligence also captured Hong’s attention because of its medical potential. He believed those specialties could benefit from long term recovery analysis through advanced technology. Despite that interest, he arrived at New York Tech without practical knowledge about artificial intelligence mechanics.
Hong’s direction changed during his first semester after meeting Associate Professor Milan Toma. Toma welcomed students interested in research through an open door approach. His expertise included artificial intelligence assisted medical diagnostics and algorithmic medicine. Hong decided to pursue his own research project after conversations with his mentor.
Artificial Intelligence Steps Into Patient Recovery
Hong completed his first medical school year with three published research articles. Associate Professor Toma praised Hong’s initiative, productivity, and intellectual curiosity throughout their collaboration. Hong achieved first author status across all three published research papers. His accomplishments came despite limited machine learning experience before medical school began.
One study became Hong’s personal favorite because of its practical medical application. The research examined thermography for automated epidural block assessment during labor. The paper appeared in *Frontiers* and focused upon women who received epidurals.
Epidural blocks can fail in up to 12% of patients during treatment. Physicians traditionally confirm effectiveness through pinprick or ice cube testing afterward. Those methods may cause additional discomfort during an already stressful clinical experience. Hong sought a less intrusive option that could reduce unnecessary patient distress.
Previous datasets showed epidural blocks increase foot temperature through blood vessel dilation. Hong recognized thermal cameras could capture that physiological response without physical stimulation. That concept offered a noninvasive and less stressful confirmation approach for patients.
The proposed artificial intelligence approach relied upon thermal imaging instead of painful confirmation methods. Automated analysis could support faster evaluation through objective temperature information from patient feet. The research explored how advanced technology might improve patient comfort during clinical care.
Innovation Moves Forward With Careful Boundaries
Hong built the system around a deep learning model called U-Net. The model functions as an automated assistant rather than an independent clinical decision maker. It instantly segments each thermal image before further analytical evaluation begins. That process helps prepare visual information for rapid medical assessment.
The system identifies patient feet within thermal images through automated image segmentation. It then extracts mean temperature values from those identified anatomical regions. Real time analysis supports efficient clinical evaluation without manual image interpretation.
Hong said the approach converts subjective patient sensations into objective visual measurements. Quantified temperature maps help reduce potential human error during clinical evaluation. Objective information can support more consistent assessment across similar medical situations. Reliable measurements may also strengthen physician confidence during routine clinical practice.
The technology supports physicians instead of replacing professional medical judgment. Automated assistance can improve workflow while clinicians retain responsibility for patient care. Human expertise remains essential whenever important medical decisions require careful evaluation.
The research emphasizes practical support instead of complete clinical automation through artificial intelligence. Physicians remain responsible for interpretation, treatment decisions, and overall patient safety. That balance reflects a careful approach toward responsible medical technology adoption.
Responsible Innovation Defines the Road Ahead
Hong believes careful evaluation should precede any artificial intelligence use inside clinical practice. Successful laboratory results alone cannot guarantee safe hospital performance for every patient. Thorough validation remains essential before healthcare providers place confidence in emerging medical technologies. Patient safety continues to guide every future decision about clinical artificial intelligence adoption.
Hong does not believe artificial intelligence has reached independent diagnostic reliability today. Physicians should continue careful oversight because professional judgment remains clinically indispensable. Responsible medical practice still requires experienced clinicians to evaluate every important patient outcome. Artificial intelligence should support healthcare professionals instead of directing clinical care without supervision.
Hong expects artificial intelligence to become as familiar as the medical stethoscope someday. That future depends upon responsible adoption supported by careful clinical evidence and trust. Medical innovation will achieve greater value when technology strengthens patient care without compromising professional responsibility.