Post Date
Jun 23 2025

SBASSE Research Heads to MICCAI 2025

We are pleased to share that two research papers from the Syed Babar Ali School of Science and Engineering (SBASSE), LUMS, have been accepted for presentation at the 28th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2025).

MICCAI 2025 will take place from September 23–27, 2025, at the Daejeon Convention Center in South Korea, a city celebrated as the “Silicon Valley of Korea” and a hub for science and innovation. MICCAI is widely regarded as a leading global platform in the domains of medical image computing and AI-assisted healthcare technologies. Ranked ‘A’ by the CORE conference rankings, this year’s conference received a record 3,677 submissions, which is 28% more than last year. Of these, 3,447 papers underwent peer review, with only 1,014 accepted, reflecting a highly competitive 29% acceptance rate.

Both accepted papers were developed as part of MS thesis projects under the supervision of Dr. Murtaza Taj, Associate Professor in the Department of Computer Science at SBASSE, LUMS. The work highlights SBASSE’s continued emphasis on mentorship, interdisciplinary collaboration, and impactful research in emerging areas of science and technology.

1. CATVis: Context-Aware Thought Visualization
This paper investigates how the cognitive process of interpreting medical images can be visualized, offering new possibilities for training and diagnostic support in clinical environments.

 

Proposed five-stage framework for reconstructing images from EEG signals. The numbered modules correspond to the key steps in the methodology.

 

 

Authors: Hamza Ahmad (Undergrad. – FCCU, Founding AI Eng. – Uplift AI), Tariq Mehmood (MSCS, LUMS), Dr. Muhammad Haroon Shakeel (Arbisoft, PhD LUMS), Dr. Murtaza Taj (LUMS)
Hamza and Tariq contributed equally to this research.

2. Localization Lens for Improving Medical Vision-Language Models
The second study introduces a localization-based framework to enhance the performance and interpretability of AI models that combine visual and textual data in medical imaging.

 

Comparison of the existing standard Med-VLM training with the proposed
training utilizing localization lens.

 

Authors: Hasan Farooq (MSCS, LUMS), Dr. Murtaza Taj (LUMS), Dr. Mehwish Nasim (University of Western Australia), Dr. Arif Mehmood (ITU)


These recognitions are a testament to the quality of research training at SBASSE, and the global relevance of the scientific questions explored by its students and faculty.
Congratulations to all authors and collaborators on this remarkable achievement.