Ultra high speed SPECT bone imaging enabled by a deep learning enhancement method: a proof of concept

Clinical performance of deep learning-enhanced ultrafast whole-body scintigraphy in patients with suspected malignancy

BACKGROUND

To evaluate the clinical performance of two deep learning methods, one utilizing real clinical pairs and the other utilizing simulated datasets, in enhancing image quality for two-dimensional (2D) fast whole-body scintigraphy (WBS).

METHODS

A total of 83 patients with suspected bone metastasis were retrospectively enrolled. All patients underwent single-photon emission computed tomography (SPECT) WBS at speeds of 20 cm/min (1x), 40 cm/min (2x), and 60 cm/min (3x). Two deep learning models were developed to generate high-quality images from real and simulated fast scans, designated 2x-real and 3x-real (images from real fast data) and 2x-simu and 3x-simu (images from simulated fast data), respectively. A 5-point Likert scale was used to evaluate the image quality of each acquisition. Accuracy, sensitivity, specificity, and the area under the curve (AUC) were used to evaluate diagnostic efficacy. Learned perceptual image patch similarity (LPIPS) and the Fréchet inception distance (FID) were used to assess image quality. Additionally, the count-level consistency of WBS was compared between the two models.

RESULTS

Subjective assessments revealed that the 1x images had the highest general image quality (Likert score: 4.40 ± 0.45). The 2x-real, 2x-simu and 3x-real, 3x-simu images demonstrated significantly better quality than the 2x and 3x images (Likert scores: 3.46 ± 0.47, 3.79 ± 0.55 vs. 2.92 ± 0.41, P < 0.0001; 2.69 ± 0.40, 2.61 ± 0.41 vs. 1.36 ± 0.51, P < 0.0001), respectively. Notably, the quality of the 2x-real images was inferior to that of the 2x-simu images (Likert scores: 3.46 ± 0.47 vs. 3.79 ± 0.55, P = 0.001). The diagnostic efficacy for the 2x-real and 2x-simu images was indistinguishable from that of the 1x images (accuracy: 81.2%, 80.7% vs. 84.3%; sensitivity: 77.27%, 77.27% vs. 87.18%; specificity: 87.18%, 84.63% vs. 87.18%. All P > 0.05), whereas the diagnostic efficacy for the 3x-real and 3x-simu was better than that for the 3x images (accuracy: 65.1%, 66.35% vs. 59.0%; sensitivity: 63.64%, 63.64% vs. 64.71%; specificity: 66.67%, 69.23% vs. 55.1%. All P < 0.05). Objectively, both the real and simulated models achieved significantly enhanced image quality from the accelerated scans in the 2x and 3x groups (FID: 0.15 ± 0.18, 0.18 ± 0.18 vs. 0.47 ± 0.34; 0.19 ± 0.23, 0.20 ± 0.22 vs. 0.98 ± 0.59.

LPIPS

0.17 ± 0.05, 0.16 ± 0.04 vs. 0.19 ± 0.05; 0.18 ± 0.05, 0.19 ± 0.05 vs. 0.23 ± 0.04. All P < 0.05). The count-level consistency with the 1x images was excellent for all four sets of model-generated images (P < 0.0001).

CONCLUSIONS

Ultrafast 2x speed (real and simulated) images achieved comparable diagnostic value to that of standardly acquired images, but the simulation algorithm does not necessarily reflect real data.

Verification of image quality improvement of low-count bone scintigraphy using deep learning

To improve image quality for low-count bone scintigraphy using deep learning and evaluate their clinical applicability. Six hundred patients (training, 500; validation, 50; evaluation, 50) were included in this study. Low-count original images (75%, 50%, 25%, 10%, and 5% counts) were generated from reference images (100% counts) using Poisson resampling. Output (DL-filtered) images were obtained after training with U-Net using reference images as teacher data. Gaussian-filtered images were generated for comparison. Peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) to the reference image were calculated to determine image quality. Artificial neural network (ANN) value, bone scan index (BSI), and number of hotspots (Hs) were computed using BONENAVI analysis to assess diagnostic performance. Accuracy of bone metastasis detection and area under the curve (AUC) were calculated. PSNR and SSIM for DL-filtered images were highest in all count percentages. BONENAVI analysis values for DL-filtered images did not differ significantly, regardless of the presence or absence of bone metastases. BONENAVI analysis values for original and Gaussian-filtered images differed significantly at ≦25% counts in patients without bone metastases. In patients with bone metastases, BSI and Hs for original and Gaussian-filtered images differed significantly at ≦10% counts, whereas ANN values did not. The accuracy of bone metastasis detection was highest for DL-filtered images in all count percentages; the AUC did not differ significantly. The deep learning method improved image quality and bone metastasis detection accuracy for low-count bone scintigraphy, suggesting its clinical applicability.

Application of Nuclear Medicine Techniques in Musculoskeletal Infection: Current Trends and Future Prospects

Nuclear medicine has become an indispensable discipline in the diagnosis and management of musculoskeletal infections. Radionuclide tests serve as a valuable diagnostic tool for patients suspected of having osteomyelitis, spondylodiscitis, or prosthetic joint infections. The choice of the most suitable imaging modality depends on various factors, including the affected area, potential extra osseous involvement, or the impact of previous bone/joint conditions. This review provides an update on the use of conventional radionuclide imaging tests and recent advancements in fusion imaging scans for the differential diagnosis of musculoskeletal infections. Furthermore, it examines the role of radionuclide scans in monitoring treatment responses and explores current trends in their application. We anticipate that this update will be of significant interest to internists, rheumatologists, radiologists, orthopedic surgeons, rehabilitation physicians, and other specialists involved in musculoskeletal pathology.

A Review on Low-Dose Emission Tomography Post-Reconstruction Denoising with Neural Network Approaches

Low-dose emission tomography (ET) plays a crucial role in medical imaging, enabling the acquisition of functional information for various biological processes while minimizing the patient dose. However, the inherent randomness in the photon counting process is a source of noise which is amplified in low-dose ET. This review article provides an overview of existing post-processing techniques, with an emphasis on deep neural network (NN) approaches. Furthermore, we explore future directions in the field of NN-based low-dose ET. This comprehensive examination sheds light on the potential of deep learning in enhancing the quality and resolution of low-dose ET images, ultimately advancing the field of medical imaging.

Artificial Intelligence for PET and SPECT Image Enhancement

Nuclear medicine imaging modalities such as PET and SPECT are confounded by high noise levels and low spatial resolution, necessitating postreconstruction image enhancement to improve their quality and quantitative accuracy. Artificial intelligence (AI) models such as convolutional neural networks, U-Nets, and generative adversarial networks have shown promising outcomes in enhancing PET and SPECT images. This review article presents a comprehensive survey of state-of-the-art AI methods for PET and SPECT image enhancement and seeks to identify emerging trends in this field. We focus on recent breakthroughs in AI-based PET and SPECT image denoising and deblurring. Supervised deep-learning models have shown great potential in reducing radiotracer dose and scan times without sacrificing image quality and diagnostic accuracy. However, the clinical utility of these methods is often limited by their need for paired clean and corrupt datasets for training. This has motivated research into unsupervised alternatives that can overcome this limitation by relying on only corrupt inputs or unpaired datasets to train models. This review highlights recently published supervised and unsupervised efforts toward AI-based PET and SPECT image enhancement. We discuss cross-scanner and cross-protocol training efforts, which can greatly enhance the clinical translatability of AI-based image enhancement tools. We also aim to address the looming question of whether the improvements in image quality generated by AI models lead to actual clinical benefit. To this end, we discuss works that have focused on task-specific objective clinical evaluation of AI models for image enhancement or incorporated clinical metrics into their loss functions to guide the image generation process. Finally, we discuss emerging research directions, which include the exploration of novel training paradigms, curation of larger task-specific datasets, and objective clinical evaluation that will enable the realization of the full translation potential of these models in the future.

The Past, Present, and Future Role of Artificial Intelligence in Ventilation/Perfusion Scintigraphy: A Systematic Review

Ventilation-perfusion (V/Q) lung scans constitute one of the oldest nuclear medicine procedures, remain one of the few studies performed in the acute setting, and are amongst the few performed in the emergency setting. V/Q studies have witnessed a long fluctuation in adoption rates in parallel to continuous advances in image processing and computer vision techniques. This review provides an overview on the status of artificial intelligence (AI) in V/Q scintigraphy. To clearly assess the past, current, and future role of AI in V/Q scans, we conducted a systematic Ovid MEDLINE(R) literature search from 1946 to August 5, 2022 in addition to a manual search. The literature was reviewed and summarized in terms of methodologies and results for the various applications of AI to V/Q scans. The PRISMA guidelines were followed. Thirty-one publications fulfilled our search criteria and were grouped into two distinct categories: (1) disease diagnosis/detection (N = 22, 71.0%) and (2) cross-modality image translation into V/Q images (N = 9, 29.0%). Studies on disease diagnosis and detection relied heavily on shallow artificial neural networks for acute pulmonary embolism (PE) diagnosis and were primarily published between the mid-1990s and early 2000s. Recent applications almost exclusively regard image translation tasks from CT to ventilation or perfusion images with modern algorithms, such as convolutional neural networks, and were published between 2019 and 2022. AI research in V/Q scintigraphy for acute PE diagnosis in the mid-90s to early 2000s yielded promising results but has since been largely neglected and thus have yet to benefit from today's state-of-the art machine-learning techniques, such as deep neural networks. Recently, the main application of AI for V/Q has shifted towards generating synthetic ventilation and perfusion images from CT. There is therefore considerable potential to expand and modernize the use of real V/Q studies with state-of-the-art deep learning approaches, especially for workflow optimization and PE detection at both acute and chronic stages. We discuss future challenges and potential directions to compensate for the lag in this domain and enhance the value of this traditional nuclear medicine scan.

Is There a Role of Artificial Intelligence in Preclinical Imaging?

This review provides an overview of the current opportunities for integrating artificial intelligence methods into the field of preclinical imaging research in nuclear medicine. The growing demand for imaging agents and therapeutics that are adapted to specific tumor phenotypes can be excellently served by the evolving multiple capabilities of molecular imaging and theranostics. However, the increasing demand for rapid development of novel, specific radioligands with minimal side effects that excel in diagnostic imaging and achieve significant therapeutic effects requires a challenging preclinical pipeline: from target identification through chemical, physical, and biological development to the conduct of clinical trials, coupled with dosimetry and various pre, interim, and post-treatment staging images to create a translational feedback loop for evaluating the efficacy of diagnostic or therapeutic ligands. In virtually all areas of this pipeline, the use of artificial intelligence and in particular deep-learning systems such as neural networks could not only address the above-mentioned challenges, but also provide insights that would not have been possible without their use. In the future, we expect that not only the clinical aspects of nuclear medicine will be supported by artificial intelligence, but that there will also be a general shift toward artificial intelligence-assisted in silico research that will address the increasingly complex nature of identifying targets for cancer patients and developing radioligands.

Low-dose sinogram restoration enabled by conditional GAN with cross-domain regularization in SPECT imaging

In order to generate high-quality single-photon emission computed tomography (SPECT) images under low-dose acquisition mode, a sinogram denoising method was studied for suppressing random oscillation and enhancing contrast in the projection domain. A conditional generative adversarial network with cross-domain regularization (CGAN-CDR) is proposed for low-dose SPECT sinogram restoration. The generator stepwise extracts multiscale sinusoidal features from a low-dose sinogram, which are then rebuilt into a restored sinogram. Long skip connections are introduced into the generator, so that the low-level features can be better shared and reused, and the spatial and angular sinogram information can be better recovered. A patch discriminator is employed to capture detailed sinusoidal features within sinogram patches; thereby, detailed features in local receptive fields can be effectively characterized. Meanwhile, a cross-domain regularization is developed in both the projection and image domains. Projection-domain regularization directly constrains the generator via penalizing the difference between generated and label sinograms. Image-domain regularization imposes a similarity constraint on the reconstructed images, which can ameliorate the issue of ill-posedness and serves as an indirect constraint on the generator. By adversarial learning, the CGAN-CDR model can achieve high-quality sinogram restoration. Finally, the preconditioned alternating projection algorithm with total variation regularization is adopted for image reconstruction. Extensive numerical experiments show that the proposed model exhibits good performance in low-dose sinogram restoration. From visual analysis, CGAN-CDR performs well in terms of noise and artifact suppression, contrast enhancement and structure preservation, particularly in low-contrast regions. From quantitative analysis, CGAN-CDR has obtained superior results in both global and local image quality metrics. From robustness analysis, CGAN-CDR can better recover the detailed bone structure of the reconstructed image for a higher-noise sinogram. This work demonstrates the feasibility and effectiveness of CGAN-CDR in low-dose SPECT sinogram restoration. CGAN-CDR can yield significant quality improvement in both projection and image domains, which enables potential applications of the proposed method in real low-dose study.

Detection of prostate cancer bone metastases with fast whole-body 99mTc-HMDP SPECT/CT using a general-purpose CZT system

BACKGROUND

We evaluated the effects of acquisition time, energy window width, and matrix size on the image quality, quantitation, and diagnostic performance of whole-body ^99mTc-HMDP SPECT/CT in the primary metastasis staging of prostate cancer.

METHODS

Thirty prostate cancer patients underwent ^99mTc-HMDP SPECT/CT from the top of the head to the mid-thigh using a Discovery NM/CT 670 CZT system with list-mode acquisition, 50-min acquisition time, 15% energy window width, and 128 × 128 matrix size. The acquired list-mode data were resampled to produce data sets with shorter acquisition times of 41, 38, 32, 26, 20, and 16 min, narrower energy windows of 10, 8, 6, and 4%, and a larger matrix size of 256 × 256. Images were qualitatively evaluated by three experienced nuclear medicine physicians and quantitatively evaluated by noise, lesion contrast and SUV measurements. Diagnostic performance was evaluated from the readings of two experienced nuclear medicine physicians in terms of patient-, region-, and lesion-level sensitivity and specificity.

RESULTS

The originally acquired images had the best qualitative image quality and lowest noise. However, the acquisition time could be reduced to 38 min, the energy window narrowed to 8%, and the matrix size increased to 256 × 256 with still acceptable qualitative image quality. Lesion contrast and SUVs were not affected by changes in acquisition parameters. Acquisition time reduction had no effect on the diagnostic performance, as sensitivity, specificity, accuracy, and area under the receiver-operating characteristic curve were not significantly different between the 50-min and reduced acquisition time images. The average patient-level sensitivities of the two readers were 88, 92, 100, and 96% for the 50-, 32-, 26-, and 16-min images, respectively, and the corresponding specificities were 78, 84, 84, and 78%. The average region-level sensitivities of the two readers were 55, 58, 59, and 56% for the 50-, 32-, 26-, and 16-min images, respectively, and the corresponding specificities were 95, 98, 96, and 95%. The number of equivocal lesions tended to increase as the acquisition time decreased.

CONCLUSION

Whole-body ^99mTc-HMDP SPECT/CT can be acquired using a general-purpose CZT system in less than 20 min without any loss in diagnostic performance in metastasis staging of high-risk prostate cancer patients.