CL102INTERMEDIATECLINICAL

AI for Medical Analysis 2: Deep Learning and Research Frontiers

Deep learning across medical modalities, modern research methods and safe deployment.

Course Information

Course Information
ModeInstructor-led
TermSP2 · Mar–Apr
Duration8 weeks over 2 months · about 32 hours
Credits1.5 credits
Counts towardStandalone course
PrerequisitesCL101, or equivalent knowledge of data preparation, evaluation, neural networks and CNNs.
Who it is forLearners who have completed CL101: engineers, medical students, residents and clinicians.
FacultyDr. Chandra Sekhar Vorugunti
AssessmentWeekly quizzes 15% · two assignments 45% · capstone 40%. 60% to pass.

About the course

Continues CL101. Learners apply deep learning across medical modalities, interpret modern research methods and assess the evidence a model needs before clinical deployment.

Learning outcomes

  • Apply deep learning to ECG and patient time series, and test explanations and external validity
  • Describe segmentation, foundation models and large language models, including their failure modes
  • Assess privacy, fairness and regulation, including the DPDP Act 2023 and ICMR guidelines
  • Explain a recent CVPR or MICCAI paper in clinical terms and judge whether it would hold in a hospital
  • Critically appraise a medical AI study with TRIPOD+AI, CLAIM and CONSORT-AI

Course modules

8 modules

  • 1-D CNNs, RNNs and LSTMs

Labs and projects: Weekly notebooks; two assignments; a capstone critique, project or reproduction.

Books and readings: Reading list of recent MICCAI, CVPR and Nature-family papers