Two patients with apparently similar cancers can respond very differently to the same treatment. Understanding why has become one of the central questions in precision oncology, and researchers are increasingly examining whether medical images can provide part of the answer.
This is an area being explored by radiologist and researcher Dr. Ohmini Krishnamurthy Rajendran, whose work brings together medical imaging, artificial intelligence, radiomics, radiogenomics and computational oncology.
Radiomics uses computational techniques to extract quantitative characteristics from medical images. Radiogenomics takes the concept further by examining relationships between imaging characteristics and the molecular or genomic features of disease. Together, these approaches could allow imaging to provide information extending beyond the location and physical dimensions of a tumour.
Dr. Rajendran’s interest in this field developed from the practical limitations she encountered through clinical radiology. Cancer decisions rarely depend upon a single source of information. Radiologists examine imaging, pathologists study tissue, molecular laboratories analyse biomarkers and genomic characteristics, while oncologists consider the patient’s overall clinical condition and response to treatment.
Her research reflects an effort to understand how these different sources of information can be analysed together.
This has led her work towards multimodal approaches that connect imaging with clinical and biological data. Such models are particularly relevant to questions involving cancer detection, prognosis, treatment-response prediction and individualised therapeutic planning.
The distinction is important. An artificial-intelligence system that identifies an abnormality on an image addresses one problem. A system capable of considering imaging patterns alongside biological and clinical information addresses a substantially broader clinical question: what does the disease mean for this particular patient?
Dr. Rajendran has continued to develop this interdisciplinary direction through peer-reviewed scholarly work spanning medical imaging and healthcare artificial intelligence. Her clinical background gives the research a practical dimension because the ultimate value of computational medicine depends not merely upon algorithmic performance but upon whether the resulting information can meaningfully assist physicians.
Her work forms part of a broader transition occurring in cancer medicine, where imaging, molecular biology and computational analysis are increasingly converging. For radiologists working within this environment, understanding the image may increasingly mean understanding the biology behind it as well.
