Development of an Image Analysis-Based Prognosis Score Using Google’s Teachable Machine in Melanoma

Forchhammer, Stephan; Abu-Ghazaleh, Amar; Metzler, Gisela; Garbe, Claus; Eigentler, Thomas

doi:10.3390/cancers14092243

Cited by 11 publications

(8 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent studies have shown that TM, a code-free DL platform, can also be successfully used to analyze medical images. [22][23][24] Previous studies reported classification accuracy rates above 90% when transfer learning was used to train relatively large datasets. Despite the high accuracy levels of DL-based models in many ophthalmic diseases, there are still several clinical and technical challenges to their real-time application in clinical practice.…”

Section: Discussionmentioning

confidence: 99%

Efficiency of web-based code-free artificial intelligence platform in classification of vitreomacular interface diseases

Kirik

2024

Eur Eye Res

View full text Add to dashboard Cite

The aim of this study is to evaluate the effectiveness of Teachable Machine (TM), a code-free web-based artificial intelligence (AI) platform, in the detection and classification of vitreomacular interface diseases (VMIDs) in optical coherence tomography (OCT) images. Methods: A dataset of 445 cross-sectional OCT images from patients with VMID, along with 200 images from healthy individuals, was retrospectively prepared at a tertiary health-care institution. The OCT images were categorized into three groups: Epiretinal membrane (ERM), macular hole (MH), and vitreomacular traction (VMT). Subsequently, a deep learning (DL) model for VMID classification was developed using TM, a code-free web-based AI platform. The model underwent training on 160 ERM, 96 MH, 100 VMT, and 160 normal images, followed by testing on 40 ERM, 25 VMT, 24 MH, and 40 normal images. Sensitivity, specificity, and receiver operating characteristic curve were calculated to evaluate the effectiveness of the developed model. Results: The DL model showed 100% sensitivity and specificity in detecting any VMID compared to normal eyes. For detecting VMT, TM had 100% sensitivity, 98.08% specificity, and an AUC of 0.99. In ERM detection, sensitivity and specificity were both 100%, with an AUC of 1.00. MH detection had 91.67% sensitivity, 100% specificity, and AUC of 0.958. Conclusion:This study demonstrates that TM can be used with high efficiency in detecting and classifying VMID. TM application, which performs image classification with DL, can be considered an effective alternative, especially for physicians who do not have coding knowledge to develop AI models.

show abstract

Section: Discussionmentioning

confidence: 99%

Efficiency of web-based code-free artificial intelligence platform in classification of vitreomacular interface diseases

Kirik

2024

Eur Eye Res

View full text Add to dashboard Cite

show abstract

“…TM was able to accurately predict prognosis for low-and high-risk groups. Researchers concluded that TM showed promise in considering melanoma risk using a slide image [36]. In one final example, TM was used to determine a tooth-marked tongue or not [37].…”

Section: Overview Of Tmmentioning

confidence: 99%

“…Scientific data are often classified into two categories. We provided three contextualized examples earlier-waterbirds or shorebirds [35], melanoma low risk or high risk [36] and tooth-marked tongue or non-tooth-marked tongue [37]. However, a binary classification system may not be the most effective for all scientific TM uses.…”

Section: Curricular Considerationsmentioning

confidence: 99%

The Impact of Teachable Machine on Middle School Teachers’ Perceptions of Science Lessons after Professional Development

Kurz,

Jayasuriya,

Swisher

et al. 2024

Education Sciences

View full text Add to dashboard Cite

Technological advances in computer vision and machine learning image and audio classification will continue to improve and evolve. Despite their prevalence, teachers feel ill-prepared to use these technologies to support their students’ learning. To address this, in-service middle school teachers participated in professional development, and middle school students participated in summer camp experiences that included the use of Google’s Teachable Machine, an easy-to-use interface for training machine learning classification models. An overview of Teachable Machine is provided. As well, lessons that highlight the use of Teachable Machine in middle school science are explained. Framed within Personal Construct Theory, an analysis of the impact of the professional development on middle school teachers’ perceptions (n = 17) of science lessons and activities is provided. Implications for future practice and future research are described.

show abstract

“…Teachable machine is a platform used to develop technology from Machine Learning [37], [38]. Machine learning itself is a technology that can be used for learning media in the introduction of fish species in this study [39]. This technology allows the user to manage and provide the ability for computers to learn without the need for traditional programs in it.…”

Section: Teachable Machinementioning

confidence: 99%

Implementation Of Android-Based Fish Detection & Recognition System Using Convolutional Neural Network Method

Faikul Umam,

Ach. Dafid

2023

Technium

View full text Add to dashboard Cite

Today the development of robots has increased a lot. Over time the robot can used as a medium of learning and education. The maze mapping system can be used as educational media in robotics. In this study, the idea was obtained to design a robot moving objects that are able to move on a flat striped plane like a line tracer robot. Robot it adopts a maze mapping system to find the fastest path in moving objects. These objects are in the form of mini objects that have been given a color consisting of red, yellow, green, and blue. The four colors are used as a reference for the robot to detect objects based on the color that will be moved from the object detection start point to the available finish point. The method to be used on the robot is A*. The A* algorithm is able to find the fastest route on the path traversed by the robot, by adding up the actual distance with the estimated distance thus making it optimal in the search for the route. The microcontroller used on This robot is an Arduino Due which functions to enter data from the results of sensor readings is on the robots. The sensors consist of an infrared module that functions as a detector line and TCS34725 color sensor to detect the object to be moved. This wheeled robot using a DC motor with a voltage of 12V to drive the two wheels. Than 10 times experiment to determine the fastest path with 1 type of color, obtained success with 70%. The results of this study can be used as a comparison material for the A* method with the A* method another fastest path search.

show abstract

Development of an Image Analysis-Based Prognosis Score Using Google’s Teachable Machine in Melanoma

Cited by 11 publications

References 33 publications

Efficiency of web-based code-free artificial intelligence platform in classification of vitreomacular interface diseases

Efficiency of web-based code-free artificial intelligence platform in classification of vitreomacular interface diseases

The Impact of Teachable Machine on Middle School Teachers’ Perceptions of Science Lessons after Professional Development

Implementation Of Android-Based Fish Detection & Recognition System Using Convolutional Neural Network Method

Contact Info

Product

Resources

About