Google Colab Research Articles

Brain Tumour Detection and Classification using Deep Convolutional Neural Network (DCNN)

Tumour is the undesired mass in the body. Brain tumour is the significant growth of brain cells. Manual method of classifying is time consuming and can be done at selective diagnostic centers only. Brain tumour classification is crucial task to do since treatment is based on different location and size of it. Magnetic Resonance Imaging (MRI) is most suitable way to do so. Hence there is a need to build such system which will automatically classify the brain tumour type based on input MR images only. The objective of the proposed system isto classify the brain tumour images into three sub-types: Meningioma, Glioma and Pituitary using convolutional neural network (CNN) and Support vector machine (SVM). Images from the dataset are downsized to reduce computation and some salt noise is added to make model robust and increase the dataset. The performance comparison is done on Google Colab and tensorflow platform in python language.

Semantic segmentation based on enhanced gated pyramid network with lightweight attention module

Semantic segmentation has made tremendous progress in recent years. The development of large datasets and the regression of convolutional models have enabled effective training of very large semantic model. Nevertheless, higher capacity indicates a higher computational problem, thus preventing real-time operation. Yet, due to the limited annotations, the models may have relied heavily on the available contexts in the training data, resulting in poor generalization to previously unseen scenes. Therefore, to resolve these issues, Enhanced Gated Pyramid network (GPNet) with Lightweight Attention Module (LAM) is proposed in this paper. GPNet is used for semantic feature extraction and GPNet is enhanced by the pre-trained dilated DetNet and Dense Connection Block (DCB). LAM approach is applied to habitually rescale the different feature channels weights. LAM module can increase the accuracy and effectiveness of the proposed methodology. The performance of proposed method is validated using Google Colab environment with different datasets such as Cityscapes, CamVid and ADE20K. The experimental results are compared with various methods like GPNet-ResNet-101 and GPNet-ResNet-50 in terms of IoU, precision, accuracy, F1 score and recall. From the overall analysis cityscapes dataset achieves 94.82% pixel accuracy.

Price Prediction System of Basic Commodities Using Long Short-Term Memory Method: Analysis and Implementation for Future Projections

The uncontrollable prices of basic commodities, especially food commodities, have resulted in losses for producers and consumers in the city of Cirebon. To be able to bridge these problems, it is necessary to make the right decisions. A prediction system is one of the elements that can be used to support the right decision-making. Predictions in decision-making are based on existing data at present and in the past so that they can be used to describe conditions in accordance with the objectives achieved. With an accurate price prediction system for staple food commodities, it is hoped that decision-making will be able to decide on good policies from the Cooperative, Small and Medium Enterprises Service for Trade and Industry (DKUKMPP) for the people of Cirebon City. The method used is Long Short-Term Memory (LSTM). The data used for the basic prices of food commodities, namely the last 3 years in 2020 - 2022, are sourced from the official website of the Ministry of Trade's Market Monitoring System and Basic Needs using Google Colab tools.

ВИЯВЛЕННЯ ОБ’ЄКТІВ НА ЗОБРАЖЕННІ В ПОТОКОВОМУ РЕЖИМІ ПРИ ВИКОРИСТАННІ YOLOv5 і FASTER R-CNN

The accuracy of the model can be one of the main indicators, on a basis of which it is possible to conclude about the suitability of the model for its practical operation. However, taking into account the specifics of the identified task, it is also worth paying attention to the speed of the model, since there is a need to process data in streaming mode. To investigate the possibilities of using machine learning in an applied problem, two groups of object recognition models considered: YOLOv5 and Faster R-CNN. The purpose of the study is to analyze the architectural solutions of the most common object detection models YOLOv5 and Faster R-CNN to build a model to improve the speed and accuracy of object detection in an applied task or further combine them. A total of 550 training images and 105 validation images collected. A dataset of 573 images from the new location also collected for final validation of the models. The use of Roboflow provided for image annotation, which allows not only to mark images, but also to export annotated data sets in various formats. Training and validation of the models carried out on the Google Colab platform. The platform uses the Python programming language and the PyTorch framework. The yolov5 and detecron2 libraries for YOLOv5 and Faster R-CNN, respectively, used for model training and validation. To determine whether the result belongs to one of the four groups, the IOU metric is used, which is the ratio of the intersection area to the area of the union of the correct and predicted bounding frames. The size of the trained YOLOv5 and Faster R-CNN models was 40.2 MB and 230.8 MB, respectively. The models tested on the second validation set. As result of the study, a set of data from video surveillance cameras collected and anno-tated using RoboFlow. The main representatives of two groups of object detection algorithms YOLOv5 and Faster R-CNN trained using the prepared data set. The results showed that both models have their advantages and disadvantages, both models are applicable for different tasks.

Use of artificial intelligence in the detection of the critical view of safety during laparoscopic cholecystectomy

Purposeto evaluate the use of artificial intelligence to detect the critical view of safety during elective laparoscopic cholecystectomy. Methodsprospective, observational study evaluating the detection of the critical view of safety with an artificial intelligence software in a consecutive series of elective laparoscopic cholecystectomies, compared to the blinded evaluation of 3 surgeons. The program was created using the digital tools PyCharm, Google Colab Pro and Ultralytics YOLO v8. Results40 consecutive elective laparoscopic cholecystectomies were included in the study. The program was able to detect the critical view of safety in all cases, in accordance to the experts blinded opinion. Conclusionin this preliminary experience, an artificial intelligence software was able to detect the critical view of safety in elective laparoscopic cholecystectomies. Its application during non elective cases, where the critical view of safety is harder to achieve, might warrant further studies.

Differentiation of invasive ductal and lobular carcinoma of the breast using MRI radiomic features: a pilot study.

Breast cancer (BC) is one of the main causes of cancer-related mortality among women. For clinical management to help patients survive longer and spend less time on treatment, early and precise cancer identification and differentiation of breast lesions are crucial. To investigate the accuracy of radiomic features (RF) extracted from dynamic contrast-enhanced Magnetic Resonance Imaging (DCE MRI) for differentiating invasive ductal carcinoma (IDC) from invasive lobular carcinoma (ILC). This is a retrospective study. The IDC of 30 and ILC of 28 patients from Dukes breast cancer MRI data set of The Cancer Imaging Archive (TCIA), were included. The RF categories such as shape based, Gray level dependence matrix (GLDM), Gray level co-occurrence matrix (GLCM), First order, Gray level run length matrix (GLRLM), Gray level size zone matrix (GLSZM), NGTDM (Neighbouring gray tone difference matrix) were extracted from the DCE-MRI sequence using a 3D slicer. The maximum relevance and minimum redundancy (mRMR) was applied using Google Colab for identifying the top fifteen relevant radiomic features. The Mann-Whitney U test was performed to identify significant RF for differentiating IDC and ILC. Receiver Operating Characteristic (ROC) curve analysis was performed to ascertain the accuracy of RF in distinguishing between IDC and ILC. Ten DCE MRI-based RFs used in our study showed a significant difference (p <0.001) between IDC and ILC. We noticed that DCE RF, such as Gray level run length matrix (GLRLM) gray level variance (sensitivity (SN) 97.21%, specificity (SP) 96.2%, area under curve (AUC) 0.998), Gray level co-occurrence matrix (GLCM) difference average (SN 95.72%, SP 96.34%, AUC 0.983), GLCM interquartile range (SN 95.24%, SP 97.31%, AUC 0.968), had the strongest ability to differentiate IDC and ILC. MRI-based RF derived from DCE sequences can be used in clinical settings to differentiate malignant lesions of the breast, such as IDC and ILC, without requiring intrusive procedures.

IoT-Based Intrusion Detection System Using New Hybrid Deep Learning Algorithm

The most significant threat that networks established in IoT may encounter is cyber attacks. The most commonly encountered attacks among these threats are DDoS attacks. After attacks, the communication traffic of the network can be disrupted, and the energy of sensor nodes can quickly deplete. Therefore, the detection of occurring attacks is of great importance. Considering numerous sensor nodes in the established network, analyzing the network traffic data through traditional methods can become impossible. Analyzing this network traffic in a big data environment is necessary. This study aims to analyze the obtained network traffic dataset in a big data environment and detect attacks in the network using a deep learning algorithm. This study is conducted using PySpark with Apache Spark in the Google Colaboratory (Colab) environment. Keras and Scikit-Learn libraries are utilized in the study. ‘CICIoT2023’ and ‘TON_IoT’ datasets are used for training and testing the model. The features in the datasets are reduced using the correlation method, ensuring the inclusion of significant features in the tests. A hybrid deep learning algorithm is designed using one-dimensional CNN and LSTM. The developed method was compared with ten machine learning and deep learning algorithms. The model’s performance was evaluated using accuracy, precision, recall, and F1 parameters. Following the study, an accuracy rate of 99.995% for binary classification and 99.96% for multiclassification is achieved in the ‘CICIoT2023’ dataset. In the ‘TON_IoT’ dataset, a binary classification success rate of 98.75% is reached.

Classification of optical coherence tomography images using deep machine learning methods

Backgraund. Optical coherence tomography (OCT) is a modern high-tech and informative method for detecting pathology of the retina and preretinal layers of the vitreous body. However, the description and interpretation of the research results require high qualifications and special training of an ophthalmologist, and significant time expenditure for the doctor and the patient. At the same time, the use of mathematical models based on artificial neural networks (ANN- models) currently makes it possible to automate many processes associated with image processing. Therefore, solving problems associated with automating the process of classifying OCT images based on ANN models is actual.&#x0D; Aims. To develop architectures of mathematical (computer) models based on deep learning of convolutional neural networks (CNN) for classification of OCT images of the retina. To compare the results of computational experiments conducted using Python tools in the Google Colaboratory with single-model and multi-model approaches and evaluate classification accuracy. To make conclusions about the optimal architecture of ANN models and the values of the hyperparameters used.&#x0D; Materials and methods. The original dataset, which was anonymized OCT images of real patients, included more than 2000 images obtained directly from the device in a resolution of 1920 × 969 × 24 BPP. The number of image classes is 12. To create the training and validation data sets, a subject area of 1100 × 550 × 24 BPP was “cut out.” Various approaches were studied: the possibility of using pretrained CNNs with transfer learning, techniques for resizing and augmenting images, as well as various combinations of hyperparameters of ANN-models. When compiling the model, the following parameters were used: Adam optimizer, categorical_crossentropy loss function, accuracy metric. All technological processes with images and ANN-models were carried out using Python language tools in Google Colaboratory.&#x0D; Results. Single-model and multi-model principles for classifying OCT images of the retina are proposed. Computational experiments on automated classification of such images obtained from a DRI OCT Triton 3D tomograph using various ANN model architectures showed an accuracy of 98-100% during training and validation and 85% during an additional test, which is a satisfactory result. The optimal architecture of the ANN model - a six-layer convolutional network - was selected and the values of its hyperparameters were determined.&#x0D; Conclusions. The results of deep training of convolutional neural network models with various architectures, their validation and testing showed satisfactory classification accuracy of retinal OCT images. These developments can be used in decision support systems in the field of ophthalmology.

Plant Disease Identification Using Machine Learning Algorithms on Single-Board Computers in IoT Environments

This paper investigates the usage of machine learning (ML) algorithms on agricultural images with the aim of extracting information regarding the health of plants. More specifically, a custom convolutional neural network is trained on Google Colab using photos of healthy and unhealthy plants. The trained models are evaluated using various single-board computers (SBCs) that demonstrate different essential characteristics. Raspberry Pi 3 and Raspberry Pi 4 are the current mainstream SBCs that use their Central Processing Units (CPUs) for processing and are used for many applications for executing ML algorithms based on popular related libraries such as TensorFlow. NVIDIA Graphic Processing Units (GPUs) have a different rationale and base the execution of ML algorithms on a GPU that uses a different architecture than a CPU. GPUs can also implement high parallelization on the Compute Unified Device Architecture (CUDA) cores. Another current approach involves using a Tensor Processing Unit (TPU) processing unit carried by the Google Coral Dev TPU Board, which is an Application-Specific Integrated Circuit (ASIC) specialized for accelerating ML algorithms such as Convolutional Neural Networks (CNNs) via the usage of TensorFlow Lite. This study experiments with all of the above-mentioned devices and executes custom CNN models with the aim of identifying plant diseases. In this respect, several evaluation metrics are used, including knowledge extraction time, CPU utilization, Random Access Memory (RAM) usage, swap memory, temperature, current milli Amperes (mA), voltage (Volts), and power consumption milli Watts (mW).

DETECTION OF CRIME SCENE OBJECTS FOR EVIDENCE ANALYSIS USING DEEP LEARNING TECHNIQUES

Research on the detection of objects at crime scenes has flourished in the last two decades. Researchers have been concentrating on colour pictures, where lighting is a crucial component, since this is one of the most pressing issues in computer vision, with applications spanning surveillance, security, medicine, and more. However, nighttime monitoring is crucial since most security problems cannot be seen by the naked eye. That's why it's crucial to record a dark scene and identify the things at a crime scene. Even when its dark out, infrared cameras are indispensable. Both military and civilian sectors will benefit from the use of such methods for nighttime navigation. On the other hand, IR photographs have issues with poor resolution, lighting effects, and other similar issues. Surveillance cameras with infrared (IR) imaging capabilities have been the focus of much study and development in recent years. This research work has attempted to offer a good model for object recognition by using IR images obtained from crime scenes using Deep Learning. The model is tested in many scenarios including a central processing unit (CPU), Google COLAB, and graphics processing unit (GPU), and its performance is also tabulated.

94 Smartphone integration in anemia detection for small ruminants with Convolutional Neural Networks

Abstract Anemia, often caused by internal parasites like Haemonchus contortus, presents significant health and productivity challenges for small ruminants. The primary goal of this study was to accurately distinguish between healthy and anemic goats using an image classification system focused on eye conjunctiva images. Over 5,000 eye conjunctiva images from 75 goats were collected at Fort Valley State University farms over a 3-mo period using smartphone cameras. These images were randomly divided into training (70%) and testing (30%) datasets, with each group containing five subfolders corresponding to FAMACHA scores of 1, 2, 3, 4, and 5. A Convolutional Neural Network (CNN) algorithm was utilized for image analysis, incorporating data augmentation techniques such as Resize, RandomHorizontalFlip, RandomVerticalFlip, and RandomRotation. The CNN model was built on the Google Colaboratory platform using CUDA 11.2 and the PyTorch machine learning framework, incorporating three ConvNet layers. The model training used the Adam Optimizer with a slower learning rate of 0.001 and a weight decay of 0.0001 to prevent exploding gradient issues, alongside ReLU and the cross-entropy loss function over 1000 epochs. Results demonstrate that the CNN model was highly effective in classifying eye conjunctiva images of goats to detect anemia based on FAMACHA scores.

Comparison of Long Short-Term Memory and Convolutional Neural Network Models for Emergency Department Patients’ Arrival Daily Forecasting

Background: One of the most critical challenges in the emergency department (ED) is overcrowding, which creates negative consequences for patients and staff. Therefore, predicting the rate of patients entering the ED can help manage resources in this department effectively. Objectives: According to the time of data collection, we intended to predict the volume of patient admissions to the ED in epidemic conditions, such as COVID-19 and non-epidemic. In addition, we planned to compare the performance of the LSTM and CNN models. Methods: The collected data consists of three main categories. The first category pertains to air pollutants, provided by the Tehran air quality control organization. The second type relates to data from the Iran Meteorological Organization, and the third category includes the number of patients admitted to the ED of a hospital in Tehran. We also incorporated binary indicators for epidemic and non-epidemic conditions in the dataset. The data collection period spans from February 2018 to March 2022. We employed the Dickey-Fuller test to assess the stationarity of the data. After preprocessing, we independently developed long short-term memory (LSTM) neural network and convolutional neural network (CNN) models, considering various time windows of previous days. Keras and Tensorflow libraries in Python, along with the Google Colab environment, were utilized to execute the models. Results: The LSTM model exhibited the lowest root mean square error (RMSE) and mean absolute error (MAE) with a time window of the last seven days, while the CNN model outperformed the LSTM model with a time window of the previous 13 days. Additionally, the CNN model required less execution time than the LSTM model. Conclusions: In conclusion, deep learning algorithms prove suitable for analyzing multivariate time series data. The CNN model demonstrated the lowest prediction error.

Методические рекомендации по реализации алгоритма шифрования при помощи кодового слова на Python

The article offers theoretical and practical material on encryption and decryption of text messages using a codeword cipher. Is shown an example of the implementation of the algorithm for cryptographic transformations of the specified cipher in the Python programming language. The development and execution of the Python program code was carried out in a free interactive cloud environment — Google Colab.

Comparative Analysis of LSTM Neural Network and SVM for USD Exchange Rate Prediction: A Study on Different Training Data Scenarios

Purpose: This paper aims to investigate and compare the performance of LSTM Neural Network and Support Vector Machines (SVM) in predicting the USD exchange rate using three different training data scenarios: 45%, 55%, and 75%. The study employs a dataset from the Indonesian Central Bureau of Statistics (BPS) for the period of January 1 to June 30, 2021, encompassing attributes USD Selling Rate.Methods: The methods involve implementing LSTM and SVM algorithms within the Python programming language using Google Colaboratory. Three distinct training data scenarios are explored to evaluate the models' robustness. Performance metrics, including Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R-squared, are employed for evaluation.Result: Results reveal that LSTM demonstrates superior prediction accuracy compared to SVM across all scenarios, even though it incurs a longer training time. Notably, in the 75% training data scenario, LSTM achieves an MAE of 49.52, RMSE of 63.08, and R-squared of 0.37906, outperforming SVM with MAE of 138.33, RMSE of 161.58, and R-squared of 0.34277.Novelty: This study innovatively compares LSTM Neural Network and Support Vector Machines (SVM) for USD exchange rate prediction across different training scenarios (45%, 55%, and 75%). Unlike previous research focusing on individual models, this study systematically evaluates both methods, highlighting the nuanced balance between prediction accuracy and training time. The findings offer novel insights into LSTM and SVM applicability in currency forecasting, providing valuable guidance for researchers and practitioners in model selection based on specific predictive task requirements.

Examples and tutorials on using Google Colab and Gradio to create online interactive student‐learning modules

AbstractThis work provides online learning modules and instructions on how educators can leverage these technologies to help students learn in a personalized online environment. In particular, we focus on Google Colab, and the features provided by the Gradio Python library to provide interactivity within these modules. The contributions of this work include: (1) Development of a teaching framework using Gradio/Colab that offers automated grading and feedback for both educators and students; (2) Design of a versatile proposal, accommodating beginners with a straightforward interface while addressing the needs of advanced learners; (3) Creation of a comprehensive set of examples tailored for teaching digital logic subjects, with adaptability for application in various computer science areas. (4) A classification of these example learning modules in terms of their learning level for the students; (5) A novel client‐server approach based on Colab/Gradio, allowing teachers to manage the main notebook efficiently while providing a lightweight and reliable interface for students. The goal of this work is to further expose educators to the remarkable capabilities that cloud computing brings to online supplemental education, noting that large language models such as ChatGPT complement this work, in that chatbots will be able to guide students in these dynamic simulations.

Machine learning tools in the tasks of selecting the determinants of socio-economic situation and development potential of Russian regions

Subject. The article addresses the selection of determinants that are significant for assessing the level of socio-economic situation and development potential of Russian regions. Objectives. The aim is to study machine learning algorithms for the selection of determinants – predictors of the level of socio-economic situation and development potential of Russian regions, to build models of classification of regions, according to the level of socio-economic situation, using various machine learning algorithms. Methods. To build classification models, I used data from the Federal State Statistics Service, the Institute of Scientific Communications, the RIA Novosti news agency, and the TAdviser Internet portal. Procedures for data classification, model parameter estimation, selection of significant determinants and visualization of results are performed, using the basic functions of the PyCaret library. Cohen's Kappa statistics and Matthews correlation coefficient were employed as priority metrics for evaluating the model productivity. The algorithms for selecting determinants are implemented in the Google Colab analytical environment. Results. I constructed multiclass classification models based on simple and ensemble machine learning algorithms. Simple classification algorithms, including logistic and ridge regression models, naive Bayesian algorithm, decision tree, support vector machine, and k-nearest neighbor methods are characterized by accuracy values at 77%, however, Cohen's Kappa statistics and the Matthews correlation coefficient only show a satisfactory relationship between the actual and predicted value of the region class. Ensemble algorithms, including random forest, gradient boosting and extreme gradient boosting, are characterized by a close relationship between the actual and forecast estimates of the classifier at a level of more than 70%. Conclusions. The random forest algorithm is recognized as the most effective classification model. The gross regional product and investments in fixed assets are informative determinants for measuring the socio-economic status.

Penerapan Algoritma Artificial Neural Network dan Economic Order Quantity dalam Memprediksi Persediaan Pengendalian BBM

Motor vehicle production in Indonesia increases every year along with increasing demand for fuel as a raw material. Generally, gas stations carry out the process of ordering fuel from Dempo on an irregular basis, the frequency of orders does not have a certain time, orders depend on sales transactions and the amount of fuel inventory available depends on the fuel in storage. Regarding prediction and control of fuel supplies, the risk at gas stations is that the volume of fuel received is different from that ordered. It is suspected that tank trucks carrying fuel during delivery from the depot to gas stations tend to experience evaporation in the tank (loses), so that the fuel quantity decreases. Requests for fuel filling are only based on monitoring without any special calculations resulting in stock being maintained and not covering consumer demand. This research is to analyze the Artificial Neural Network algorithm in predicting fuel, and determine inventory control using Economic Order Quantity. The research was conducted using data from November 2020 - October 2023. The data was processed using the ANN algorithm using Google Colab, and continued with EOQ using Microsoft Excel. The ANN parameters are 1 hidden layer with 100 units, Adam optimizer, learning rate 0.001, batch size 8 and epoch 200. Pertalite ANN test results are MSE 248852593.81 and MAE 12749.45, while Pertamax Turbo MSE 803842.94 and MAE 672, 74 provides predictions for November and December of 11,1436.82 L and 11,1960.83 L and Pertamax Turbo of 3,782.46 L and 3,660.70 L. Furthermore, in 2023 the fuel EOQ of Pertalite and Pertamax Turbo will be 8,445 L and 5,261 L, Safety Stock 3,516 L and 1,064 L, Maximum Inventory 6,042 L and 5,153 L, Re order point 2,403 L and 108 L, Order frequency 149 times and 6 times with Total Inventory Cost Rp. 178,830,302 and Rp. 7,700,459.

INTRUSION DETECTION MODEL BASED ON IMPROVED TRANSFORMER

The object of the study is the process of identifying the state of a computer network. The subject of the study are the methods of identifying the state of computer networks. The purpose of the paper is to improve the efficacy of intrusion detection in computer networks by developing a method based on transformer models. The results obtained. The work analyzes traditional machine learning algorithms, deep learning methods and considers the advantages of using transformer models. A method for detecting intrusions in computer networks is proposed. This method differs from known approaches by utilizing the Vision Transformer for Small-size Datasets (ViTSD) deep learning algorithm. The method incorporates procedures to reduce the correlation of input data and transform data into a specific format required for model operations. The developed methods are implemented using Python and the GOOGLE COLAB cloud service with Jupyter Notebook. Conclusions. Experiments confirmed the efficiency of the proposed method. The use of the developed method based on the ViTSD algorithm and the data preprocessing procedure increases the model's accuracy to 98.7%. This makes it possible to recommend it for practical use, in order to improve the accuracy of identifying the state of a computer system.

Sentiment Analysis about Anti-LGBT Campaign using the Naïve Bayes Classifier

Social media is growing so that the news that is discussed is also very fast to be known by everyone. The news or topic that is being discussed on social media is the anti-LGBT campaign. The conversation about the anti-LGBT campaign is expressed in the form of opinions that contain positive and negative feelings. The opinion is conveyed through Twitter. Twitter is a microblogging social media that allows users to create short messages and share them easily and quickly. Opinions on Twitter are used to see whether the opinion rejects or supports the anti-LGBT campaign. The use of sentiment analysis helps to see the opinion supports or rejects the anti-LGBT campaign. The algorithm used to perform sentiment analysis is the Naïve Bayes Classifier. The purpose of this study is to determine the sentiment analysis of anti-LGBT campaign tweets on Twitter. In this study using Google Colaboratory tools. The dataset used is 3103 tweets with 80% training data and 20% test data. The sentiment analysis results obtained in this study show taht Twitter users in Indonesia give more positive opinions. The use of the Naïve Bayes Classifier algorithm produces an accuracy of 68,75%, precision of 99,6%, and recall of 92,8%.

Performance evaluation of a deep learning model for automatic detection and localization of idiopathic osteosclerosis on dental panoramic radiographs

Idiopathic osteosclerosis (IO) are focal radiopacities of unknown etiology observed in the jaws. These radiopacities are incidentally detected on dental panoramic radiographs taken for other reasons. In this study, we investigated the performance of a deep learning model in detecting IO using a small dataset of dental panoramic radiographs with varying contrasts and features. Two radiologists collected 175 IO-diagnosed dental panoramic radiographs from the dental school database. The dataset size is limited due to the rarity of IO, with its incidence in the Turkish population reported as 2.7% in studies. To overcome this limitation, data augmentation was performed by horizontally flipping the images, resulting in an augmented dataset of 350 panoramic radiographs. The images were annotated by two radiologists and divided into approximately 70% for training (245 radiographs), 15% for validation (53 radiographs), and 15% for testing (52 radiographs). The study employing the YOLOv5 deep learning model evaluated the results using precision, recall, F1-score, mAP (mean Average Precision), and average inference time score metrics. The training and testing processes were conducted on the Google Colab Pro virtual machine. The test process's performance criteria were obtained with a precision value of 0.981, a recall value of 0.929, an F1-score value of 0.954, and an average inference time of 25.4 ms. Although radiographs diagnosed with IO have a small dataset and exhibit different contrasts and features, it has been observed that the deep learning model provides high detection speed, accuracy, and localization results. The automatic identification of IO lesions using artificial intelligence algorithms, with high success rates, can contribute to the clinical workflow of dentists by preventing unnecessary biopsy procedure.