KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 31

Some of the metrics are blocked by your
consent settings
Comparative performance of deep learning models and non-dermatologists in diagnosing psoriasis, dermatophytosis, and eczema
(2026-12-01)
Yodrabum, Nutcha
;
Wongpraparut, Chanisada
;
Titijaroonroj, Taravichet
;
Chularojanamontri, Leena
;
Bunyaratavej, Sumanas
Accurately differentiating scaly erythematous rashes among psoriasis, eczema, and dermatophytosis remains a clinical challenge, particularly for non-dermatologists. This study aimed to develop and evaluate deep learning models using macroscopic clinical images to classify these conditions and compare their performance with that of non-specialists. A total of 2940 images were sourced from public datasets, the Siriraj Dermatology databank, and newly collected images from Thai participants. Among sixteen evaluated models, the Swin demonstrated the best performance and interpretability. Gradient-weighted Class Activation Mapping (Grad-CAM) visualizations confirmed that the model focused on clinically relevant lesion features. Most importantly, in a pilot comparison, the Swin outperformed non-specialists in diagnostic accuracy. However, given the limited sample size of 30 images and 30 evaluators, these results should be interpreted as exploratory. Future studies with larger datasets and diverse clinician cohorts are warranted to confirm these findings and to support clinical integration.
Some of the metrics are blocked by your
consent settings
Enhancing HR Support in a Thai Organization with LLM-Based Question Answering
(2026-01-01)
Taemkaeo, Chinnatip
;
Saetia, Chanatip
;
Chalothorn, Tawunrat
;
Titijaroonroj, Taravichet
Large language models (LLMs) enhanced with retrieval-augmented generation (RAG) and multimodal inputs are increasingly used as interfaces to organizational knowledge. However, their effectiveness in specialized, non-English enterprise settings-such as Thai HR support-remains largely unclear. In many Thai organizations, employees frequently ask detailed HR-related questions, but the relevant information is scattered across internal webpages, PDF manuals, announcements, and images, making it difficult for generic LLMs to provide accurate, policy-consistent responses. To address this issue, we develop a multimodal RAG pipeline that combines hybrid dense-sparse retrieval over a vector database and evaluate six LLM models on a private Thai Visual Question Answering (VQA) HR dataset consisting of 226 questions and reference images across five HR topics. The results show that recent multimodal models, especially Qwen2.5-VL, achieve the best performance, with the highest averages in correctness (0.54), relevance (0.75), and helpfulness (0.64), clearly outperforming older vision-language systems and a text-only reasoning model. For large-scale answer evaluation, we apply an LLM-as-a-judge approach using GPT-4.1 and Gemini 2.5 Flash. We found that it serves as a generally reliable, though imperfect, substitute for human evaluation.
Some of the metrics are blocked by your
consent settings
SkinHRNet: A Deep Learning Framework for Non-Contact Heart Rate Estimation and Arterial–Venous Sufficiency Status Classification from Short Skin Videos
(2026-01-01)
Traivinidsreesuk, Chetsadaporn
;
Yodrabum, Nutcha
;
Winaikosol, Kengkart
;
Chaikangwan, Irin
;
Prompattanapakdee, Jiraya
Traditional methods for monitoring flap health in reconstructive surgery are often invasive and rely on subjective assessment. This study addresses clinically motivated monitoring challenges by evaluating two tasks using a dataset of 1,018 short skin videos: average heart rate (HR) estimation under arterial–venous sufficiency conditions and arterial–venous sufficiency status classification across sufficiency and simulated insufficiency conditions. To address these challenges, we propose SkinHRNet, a deep learning–based approach for average HR estimation and arterial–venous sufficiency status classification from short skin videos. This work contributes a short-skin-video framework that combines HR-related signal estimation with arterial–venous sufficiency classification to support the two evaluated tasks under the controlled conditions considered in this study. For average HR estimation under arterial–venous sufficiency conditions, SkinHRNet achieved a mean absolute error (MAE) of 8.66 ± 4.85 BPM. For arterial–venous sufficiency status classification, it achieved an accuracy of 0.969 ± 0.02 across the evaluated sufficiency and simulated insufficiency conditions. These findings indicate that SkinHRNet may serve as an initial research prototype for further investigation of short-video-based non-contact assessment under controlled arterial–venous sufficiency and simulated insufficiency conditions.
Some of the metrics are blocked by your
consent settings
Deep learning-based classification of lymphedema and other lower limb edema diseases using clinical images
(2025-12-01)
Lewsirirat, Thanat
;
Titijaroonroj, Taravichet
;
Apichonbancha, Sirin
;
Uthatham, Ason
;
Suwanruangsri, Veera
Lymphedema is a chronic condition characterized by lymphatic fluid accumulation, primarily affecting the limbs. Its diagnosis is challenging due to symptom overlap with conditions like chronic venous insufficiency (CVI), deep vein thrombosis (DVT), and systemic diseases, often leading to diagnostic delays that can extend up to ten years. These delays negatively impact patient outcomes and burden healthcare systems. Conventional diagnostic methods rely heavily on clinical expertise, which may fail to distinguish subtle variations between these conditions. This study investigates the application of artificial intelligence (AI), specifically deep learning, to improve diagnostic accuracy for lower limb edema. A dataset of 1622 clinical images was used to train sixteen convolutional neural networks (CNNs) and transformer-based models, including EfficientNetV2, which achieved the highest accuracy of 78.6%. Grad-CAM analyses enhanced model interpretability, highlighting clinically relevant features such as swelling and hyperpigmentation. The AI system consistently outperformed human evaluators, whose diagnostic accuracy plateaued at 62.7%. The findings underscore the transformative potential of AI as a diagnostic tool, particularly in distinguishing conditions with overlapping clinical presentations. By integrating AI with clinical workflows, healthcare systems can reduce diagnostic delays, enhance accuracy, and alleviate the burden on medical professionals. While promising, the study acknowledges limitations, such as dataset diversity and the controlled evaluation environment, which necessitate further validation in real-world settings. This research highlights the potential of AI-driven diagnostics to revolutionize lymphedema care, bridging gaps in conventional methods and supporting healthcare professionals in delivering more precise and timely interventions. Future work should focus on external validation and hybrid systems integrating AI and clinical expertise for comprehensive diagnostic solutions.
Some of the metrics are blocked by your
consent settings
Automated Verification of English Proficiency Test Scores Using OCR for Graduation Qualification
(2025-01-01)
Titijaroonroj, Taravichet
;
Maliwan, Thitiwut
;
Jaimetha, Nattakamon
;
Wattanacheep, Bhattarabhorn
An algorithm is proposed to verify English proficiency scores using OCR for graduation qualification, aiming to reduce errors and workload in the current manual verification process of officer. Score reports in image or PDF format, including TOEIC and KMITL-TEP, are processed using two deep learning-based document classifiers that identify the test type and score type prior to OCR execution. Preprocessing techniques such as noise reduction, contrast enhancement, and skew correction are applied to improve OCR accuracy. Three OCR models-Tesseract, TrOCR, and EasyOCR-are evaluated for text extraction performance. The extracted textual data are then converted into structured JSON, enabling automated rule-based comparison against graduation criteria. Evaluation performance is measured using Accuracy, F1-Score, Character Error Rate (CER), and Word Error Rate (WER). Experimental results show that the proposed system achieves 99% accuracy, demonstrating both high reliability and adaptability to institutional scoring standards. The integration of OCR significantly reduces processing time while maintaining flexibility to accommodate future changes in assessment policies.
Some of the metrics are blocked by your
consent settings
LymphoNet: A Deep Learning for Lymph Node Detection from Histological Image
(2024-01-01)
Uthatham, Ason
;
Yodrabum, Nutcha
;
Sinmaroeng, Chanya
;
Chaikangwan, Irin
;
Titijaroonroj, Taravichet
Identifying lymph nodes within a lymph node flap is crucial for precise lymph node quantification. Even when observed under a microscope, this task is known for being extremely challenging and susceptible to misidentification. Histopathology is the most reliable method for detecting lymph nodes in histopathological slides, but it is a very time-consuming and labor-intensive technique. In particular, the anatomical intricacy and significant clinical implications of the submental lymph node flap model require precise identification to ensure an effective count. Emerging deep learning techniques have shown promising capabilities for automating such meticulous tasks, potentially enhancing diagnostic efficacy and accuracy. This paper proposed LymphoNet, a deep learning model for automated detection of submental lymph nodes in histopathological slides, aiming to enhance diagnostics and reduce labor. We compared LymphoNet's performance with other models. LymphoNet demonstrated the performance, accurately identifying lymph nodes with high precision and recall, and achieving a strong F1 score. It effectively identified lymph node regions, minimizing false positives, as evidenced by a low Mean Absolute Error with non-lymph node tissues. This accuracy is necessary for lymph node flap studies in lymphedema treatment, promising to accelerate histological analysis and support pathologists and anatomists. In conclusion, LymphoNet represents a significant advancement in histopathological examination, offering precise lymph node detection that could become an essential tool for surgical planning in lymphedema management, enhancing study efficiency and treatment outcomes.
Some of the metrics are blocked by your
consent settings
Skin Video-based Blood Pressure Approximation Using CHROM with LSTM-NN
(2023-01-01)
Lumyong, Chayanin
;
Yodrabum, Nutcha
;
Winaikosol, Kengkart
;
Titijaroonroj, Taravichet
The measurement of blood pressure (BP) is an essential step in clinical practice. It is used to determine the patient's BP, which reflects the condition of the patient. Recently, there is a solution for extracting, non-invasively and with no contact, a blood pressure indicator from electrical signal like Photoplethysmography (PPG), called remote-Photoplethysmography (rPPG). This rPPG signal can be used to estimate from a video clip several vital physiological indicators for humans, especially, systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP). This paper proposed a computer method for blood pressure approximation from an input video. A chrominance method, or CHROM, was used to extract rPPG signal from a given video before forwarding it to estimate SBP and DBP values by LSTM-NN. Afterwards, MAP value was determined from SBP and DBP values by a weighting score technique. Experimental results showed that CHROM achieved the lowest mean absolute error (MAE) at 14.04, 8.37, and 9.78 for the SBP, DBP, and MAP, respectively, when compared among NN, RNN, and GRU.
Some of the metrics are blocked by your
consent settings
Chatbot-Powered Orchid Leaf Disease Classification and Management for Improved Farming
(2023-01-01)
Santisarn, Keerati
;
Titijaroonroj, Taravichet
Orchids are important economic plants in Thailand. However, orchid diseases are one of the obstacles affecting the quantity and quality of orchid yields. Classifying orchid leaf diseases requires expertise and might be difficult to follow and manage once discovered. In this research, our objective is to develop an orchid leaf disease classification model and integrate it into the chatbot with management features such as notifying and tracking when a disease is discovered. A dataset was collected from a studying orchid farm in Nakhon Pathom, Thailand, and categorized into two classes: normal orchid leaves and orchid leaves infected with yellow leaf spot disease. Well-known CNN architectures, such as DenseNet201, GoogLeNet, MobileNetV2, etc., are used to develop classification models and measure accuracy. The results show that using MobileNetV2 achieves the highest accuracy of 98%. Subsequently, the orchid leaf disease chatbot system was designed and implemented on the LINE platform.
Some of the metrics are blocked by your
consent settings
Spatial-Frequency Redistribution-based Saliency Region Detection for Thai Text Localisation
(2022-03-01)
Titijaroonroj, Taravichet
;
Suttapakti, Ungsumalee
;
Nunsong, Walairach
Saliency region detection plays an important role in computer vision applications and related areas, such as human fixation, figure-ground separation, face detection, and image compression. Nevertheless, state-of-the-art methods can moderately detect the saliency regions of a psychological pattern dataset. Therefore, this paper proposes a spatial-frequency redistribution (SFR) method to improve the efficiency of the detection of the saliency regions. The proposed SFR method consists of two main procedures: (i) adaptive cosine image construction-and-implantation and (ii) saliency region detection using complete multiple-object-implanted images. The former procedure constructs an adaptive cosine image by using a cosine function based on an object structure and then individually implants it into each object detectable. This stage provides the first significant property, spatial redistribution, to the object implanted. The adaptive cosine image redistributes the objects for controllability. Then, the latter procedure transforms a complete multiple object-implanted image into the frequency domain. At this point, the second significant property, frequency redistribution, provides the simple technique for identifying and separating the target object from the unwanted objects and background. In this paper, these properties are referred to as spatial-frequency redistribution. This method was realized as a computer program, and then such program was tested with a psychological pattern dataset and a Thai text dataset. The experimental results, when compared with state-of-the-art methods, show that the proposed SFR method can achieve the clear detection of the attention region in the psychological pattern dataset. Moreover, the proposed method can achieve a F-value of 80.28% with a recall and precision of 76.32% and 85.23%, respectively, for Thai text localisation in the Thai text dataset.
Some of the metrics are blocked by your
consent settings
Drugtionary: Drug Pill Image Detection and Recognition Based on Deep Learning
(2022-01-01)
Pornbunruang, Naphat
;
Tanjantuk, Veerapong
;
Titijaroonroj, Taravichet
Drugtionary, which is a mobile application, is developed to support people who lack medical understanding and avoid taking the wrong drug. It consists of four main features including (i) sign-up, (ii) managing profile and medication history, (iii) viewing medication information, and (iv) managing the schedule. For viewing medication information, there are three ways to retrieve the drug information–(i) text search, (ii) chatbot, and image search. We use string search and DialogFlow for text search and chatbot, respectively, whereas deep learning technique for image detection and recognition is used to search the given drug pill image. The experimental result shows that the model generated from the CenterNet method is suitable when compared to the Faster-RCNN, RetinaNet, Yolo, and SSD on our drug pill dataset. Moreover, our application is constructed by using React and React Native technology. All data are stored in the MongoDB database.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results