This paper investigates the classification of radiographic images with eleven convolutional neural network (CNN) architectures (GoogleNet, VGG-19, AlexNet, SqueezeNet, ResNet-18, Inception-v3, ResNet-50, VGG-16, ResNet-101, DenseNet-201 and Inception-ResNet-v2). The CNNs were used to classify a series of wrist radiographs from the Stanford Musculoskeletal Radiographs (MURA) dataset into two classes—normal and abnormal. The architectures were compared for different hyper-parameters against accuracy and Cohen’s kappa coefficient. The best two results were then explored with data augmentation. Without the use of augmentation, the best results were provided by Inception-ResNet-v2 (Mean accuracy = 0.723, Mean kappa = 0.506). These were significantly improved with augmentation to Inception-ResNet-v2 (Mean accuracy = 0.857, Mean kappa = 0.703). Finally, Class Activation Mapping was applied to interpret activation of the network against the location of an anomaly in the radiographs.

This paper investigates the classification of radiographic images with eleven convolutional neural network (CNN) architectures ( GoogleNet, VGG-19, AlexNet, SqueezeNet, ResNet-18, Inception-v3, ResNet-50, VGG-16, ResNet-101, DenseNet-201 and Inception-ResNet-v2 ). The CNNs were used to classify a series of wrist radiographs from the Stanford Musculoskeletal Radiographs (MURA) dataset into two classes - normal and abnormal. The architectures were compared for different hyper-parameters against accuracy and Cohen’s kappa coefficient. The best two results were then explored with data augmentation. Without the use of augmentation, the best results were provided by Inception-Resnet-v2 (Mean accuracy = 0.723 , Mean kappa = 0.506 ). These were significantly improved with augmentation to Inception-Resnet-v2 (Mean accuracy = 0.857 , Mean kappa = 0.703 ). Finally, Class Activation Mapping was applied to interpret activation of the network against the location of an anomaly in the radiographs.

Fractures of the wrist are common in Emergency Departments, where some patients are treated with a procedure called Manipulation under Anaesthesia. In some cases, this procedure is unsuccessful and patients need to revisit the hospital where they undergo surgery to treat the fracture. This work describes a geometric semi-automatic image analysis algorithm to analyse and compare the x-rays of healthy controls and patients with dorsally displaced wrist fractures (Colles’ fractures) who were treated with Manipulation under Anaesthesia. A series of 161 posterior-anterior radiographs from healthy controls and patients with Colles’ fractures were acquired and analysed. The patients’ group was further subdivided according to the outcome of the procedure (successful/unsuccessful) and pre- or post-intervention creating five groups in total (healthy, pre-successful, pre-unsuccessful, post-successful, post-unsuccessful). The semi-automatic analysis consisted of manual location of three landmarks (finger, lunate and radial styloid) and automatic processing to generate 32 geometric and texture measurements, which may be related to conditions such as osteoporosis and swelling of the wrist. Statistical differences were found between patients and controls, as well as between pre- and post-intervention, but not between the procedures. The most distinct measurements were those of texture. Although the study includes a relatively low number of cases and measurements, the statistical differences are encouraging.

Fractures of the wrist are common in Emergency Departments, where some patients are treated with a procedure called Manipulation under Anaesthesia. In some cases this procedure is unsuccessful and patients need to visit the hospital again where they undergo surgery to treat the fracture. This work describes a geometric semi-automatic image analysis algorithm to analyse and compare the x-rays of healthy controls and patients with dorsally displaced wrist fractures (Colles’ fractures) who were treated with Manipulation under Anaesthesia. A series of 161 posterior-anterior radiographs from healthy controls and patients with Colles’ fractures were acquired and analysed. The patients’ group was further subdivided according to the outcome of the procedure (successful/unsuccessful) and pre- or post-intervention creating five groups in total (healthy, pre-successful, pre-unsuccessful, post-successful, post-unsuccessful). The semi-automatic analysis consisted of manual location of three landmarks (finger, lunate and radial styloid) and automatic processing to generate 32 geometric and texture measurements, which may be related to conditions such as osteoporosis and swelling of the wrist. Statistical differences were found between patients and controls and pre- and post-intervention, but not between the procedures. The most distinct measurements were those of texture. Although the study includes a relatively low number of cases and measurements, the statistical differences are encouraging.

Bahasa adalah sistem simbolik dari suara acak, yang digunakan oleh orang untuk berkomunikasi. Mandarin berada di peringkat pertama bahasa yang paling banyak digunakan menurut survei Nationalencyklopedin, dan Mandarin menggunakan huruf yang disebut Hanzi. Manusia memiliki berbagai cara berkomunikasi, salah satunya dengan menulis. Perangkat lunak ini dapat mengenali tulisan melalui gambar dengan menerapkan Pemrosesan Gambar Digital. Penelitian ini bertujuan untuk membangun aplikasi yang dapat mendeteksi Hanzi Mandarin dengan menerapkan Digital Image Processing. Digital Image Processing mengacu pada pemrosesan gambar dua dimensi menggunakan komputer. Metode pemrosesan gambar dengan Stroke Extraction dan Template Mathcing diterapkan solusi untuk masalah ini. Stroke Extraction digunakan untuk melakukan ekstraksi stroke dasar (pi hua) pada gambar Mandarin Hanzi dan kemudian mengenalinya menggunakan Template Mathcing. Aplikasi diuji dengan Receiver Operating Characteristic sehingga kurva yang dapat membantu proses analisis dapat ditarik. Berdasarkan hasil kurva ROC yang dijelaskan, metode Stroke Extraction masih membutuhkan pengembangan karena tingkat keberhasilan yang rendah dalam menghitung sejumlah stroke. Aplikasi yang dibangun berhasil mengekstraksi Mandarin Hanzi baik tulisan tangan dan dicetak dan diterjemahkan ke dalam bahasa Indonesia dengan akurasi 76,47%. Dapat disimpulkan bahwa metode Stroke Extraction dan Template Matching cukup efisien dalam mengolah gambar digital Mandarin Hanzi untuk diterjemahkan ke dalam bahasa Indonesia.

Bahasa merupakan sistem lambang bunyi arbitrer, yang digunakan oleh masyarakat untuk berkomunikasi. Bahasa Mandarin menduduki peringkat pertama bahasa yang paling banyak digunakan menurut survei dari Nationalencyklopedin, dan bahasa Mandarin menggunakan huruf yang disebut dengan Hanzi. Manusia memiliki berbagai cara dalam berkomunikasi, salah satunya dengan tulisan. Perangkat lunak dapat mengenali tulisan melalui gambar dengan menerapkan Pengolahan Citra Digital. Penelitian ini bertujuan untuk membangun aplikasi yang dapat mendeteksi Hanzi Mandarin dengan menerapkan Pengolahan Citra Digital. Pengolahan Citra Digital merujuk pada pemrosesan gambar dua dimensi dengan menggunakan komputer. Metode Pengolahan Citra dengan Stroke Extraction dan Template Matching diimplementasikan sebagai solusi masalah. Stroke Extraction digunakan untuk melakukan ekstraksi goresan dasar (pi hua) pada citra Hanzi Mandarin dan kemudian dikenali dengan menggunakan Template Matching. Aplikasi diuji dengan Receiver Operating Characteristic sehingga didapat digambarkan sebuah kurva yang dapat membantu proses analisis. Berdasarkan kurva ROC yang tergambarkan metode Stroke Extraction masih perlu pengembangan karena rendahnya tingkat keberhasilan dalam menghitung goresan. Aplikasi yang dibangun berhasil melakukan ekstraksi Hanzi Mandarin tulisan tangan maupun cetak, dan diterjemahkan ke bahasa Indonesia dengan tingkat akurasi 76,47%. Disimpulkan bahwa metode Stroke Extraction dan Template Matching cukup efisien dalam memproses citra digital Hanzi Mandarin untuk diterjemahkan ke tulisan bahasa Indonesia.