Projects

2022 ~ 2026

Different Project using Public Data like - CT, MRI, PET, SPECT, ultrasound, mammography. 

Projects summary will be upload here soon.

2023- 2025. 

Pathology Whole Slide Cancer Tissue Image Analysis, Gastric Cancer, Liver Cancer, Breast Cancer, Colon Cancer. 

Data from 5 Korean Hospitals. 

Project 1: Lung adenocarcinoma is the most lethal subtype of lung cancer. Yet, anatomical staging alone does not always explain survival differences among patients within the same stage, especially in Stage I disease. This study assessed whether pathology-based models offer prognostic information beyond conventional staging and whether their internal classification performance is reflected in external survival prediction. Methods and Materials: We evaluated four pathology foundation models, UNI, UNI2, Phikon-v2, and CONCH, using the CPTAC-LUAD cohort for tumour differentiation classification and the TCGA-LUAD cohort for external survival validation. The model's effectiveness was evaluated using classification area under the curve (AUC), Kaplan Meier survival analysis, log-rank tests, hazard ratios (HRs), concordance index (C-index), and time-dependent AUC. On internal classification, the vision-only models performed best, with UNI2 achieving the highest AUC 82 of 0.826. This pattern was not observed in the external survival analysis. CONCH, despite having the lowest internal classification AUC (0.783), was the only model that significantly stratified overall survival in TCGA-LUAD (log84 rank p=0.0064; HR=1.495). After adjustment for age, sex, and AJCC stage, CONCH also produced the highest C85 index (0.7460). Its prognostic value was most evident in Stage I patients, where it achieved significant survival separation (p=0.031) and a three-year time-dependent AUC of 0.753. Discussion: These results suggest that vision-language pretraining may capture histological features with stronger cross-cohort prognostic relevance than vision-only training. More broadly, our findings show that internal classification performance alone is insufficient to judge clinical usefulness; pathology AI models should also be externally validated for survival before clinical deployment. 

This Project under submission Now in British Journal of Biomedical Science, Impact Factor 11.3

Project 2 ~ Project 5: Loading Soon

2020-2022. Domesticated Dog & Pet Activity Analysis. 

Dataset from Ujra Company, Seoul, Korea.

The health, safety, and well-being of household pets such as cats has become a challenging task in previous years. To estimate a cat’s behavior, objective observations of both the frequency and variability of specific behavior traits are required, which might be difficult to come by in a cat’s ordinary life. There is very little research on cat activity and cat disease analysis based on real-time data. Although previous studies have made progress, several key questions still need addressing: What types of data are best suited for accurately detecting activity patterns? Where should sensors be strategically placed to ensure precise data collection, and how can the system be effectively automated for seamless operation? This study addresses these questions by pointing out whether the cat should be equipped with a sensor, and how the activity detection system can be automated. Magnetic, motion, vision, audio, and location sensors are among the sensors used in the machine learning experiment. In this study, we collect data using three types of differentiable and realistic wearable sensors, namely, an accelerometer, a gyroscope, and a magnetometer. Therefore, this study aims to employ cat activity detection techniques to combine data from acceleration, motion, and magnetic sensors, such as accelerometers, gyroscopes, and magnetometers, respectively, to recognize routine cat activity. Data collecting, data processing, data fusion, and artificial intelligence approaches are all part of the system established in this study. We focus on One-Dimensional Convolutional Neural Networks (1D-CNNs) in our research, to recognize cat activity modeling for detection and classification. Such 1D-CNNs have recently emerged as a cutting-edge approach for signal processing-based systems such as sensor-based pet and human health monitoring systems, anomaly identification in manufacturing, and in other areas. Our study culminates in the development of an automated system for robust pet (cat) activity analysis using artificial intelligence techniques, featuring a 1D-CNN-based approach. In this experimental research, the 1D-CNN approach is evaluated using training and validation sets. The approach achieved a satisfactory accuracy of 98.9% while detecting the activity useful for cat well-being. 

Published Paper Link: https://doi.org/10.3390/s24237436

2021-2022. Anomaly Detection for CNC & CMT Machines.

Three different projects from Korean companies. 

Computer numerical control (CNC) and machine center (MCT) machines are mechanical devices that manipulate different tools using computer programming as inputs. Predicting failures in CNC and MCT machines before their actual failure time is crucial to reduce maintenance costs and increase productivity. This study is centered around a novel deep learning-based model using a 1D convolutional neural network (CNN) for early fault detection in MCT machines. We collected sensor-based data from CNC/MCT machines and applied various preprocessing techniques to prepare the dataset. Our experimental results demonstrate that the 1D-CNN model achieves a higher accuracy of 91.57% compared to traditional machine learning classifiers and other deep learning models, including Random Forest (RF) at 89.71%, multi-layer perceptron (MLP) at 87.45%, XGBoost at 89.67%, logistic regression (LR) at 75.93%, support vector machine (SVM) at 75.96%, K-nearest neighbors (KNN) at 82.93%, decision tree at 88.36%, naïve Bayes at 68.31%, long short-term memory (LSTM) at 90.80%, and a hybrid 1D CNN + LSTM model at 88.51%. Moreover, our proposed 1D CNN model outperformed all other mentioned models in precision, recall, and F-1 scores, with 91.87%, 91.57%, and 91.63%, respectively. These findings highlight the efficacy of the 1D CNN model in providing optimal performance with an MCT machine’s dataset, making it particularly suitable for small manufacturing companies seeking to automate early fault detection and classification in CNC and MCT machines. This approach enhances productivity and aids in proactive maintenance and safety measures, demonstrating its potential to revolutionize the manufacturing industry. 

Published Paper Link: https://doi.org/10.7717/peerj-cs.2389

Design & Development of IOT and AI Based Home Automation System Using Arduino

Project Study Period: 2017

Today we are living in 21st century where crime become increasing and everyone wants to secure they asset at their home. In that situation user must have system with advance technology so person do not worry when getting away from his home. It is therefore the purpose of this design to provide home security device, which send fast information to user GSM (Global System for Mobile) mobile device using SMS (Short Messaging System) and also activate - deactivate system by SMS. The Modular design of this Home Security System make expandable their capability by add more sensors on that system. Hardware of this system has been designed using microcontroller AT Mega 328, PIR (Passive Infra Red) motion sensor as the primary sensor for motion detection, camera for capturing images, GSM module for sending and receiving SMS and buzzer for alarm. For software this system using Arduino IDE for Arduino and Putty for testing connection programming in GSM module, as well as machine learning techniques were applied. This Home Security System can monitor home area that surrounding by PIR sensor and sending SMS, save images capture by camera, and make people panic by turn on the buzzer when trespassing surrounding area that detected by PIR sensor. The Modular Home Security System has been tested and succeed detect human movement.

Published Bangladeshi Domestic Conference.

2015~2019