Expert human pathologists typically require around 30 minutes to diagnose brain tumors from tissue samples extracted during surgery. A new artificially intelligent system can do it in less than 150 seconds—and it does so more accurately than its human counterparts.

New research published today in Nature Medicine describes a novel diagnostic technique that leverages the power of artificial intelligence with an advanced optical imaging technique. The system can perform rapid and accurate diagnoses of brain tumors in practically real time, while the patient is still on the operating table. In tests, the AI made diagnoses that were slightly more accurate than those made by human pathologists and in a fraction of the time. Excitingly, the new system could be used in settings where expert neurologists aren’t available, and it holds promise as a technique that could diagnose other forms of cancer as well.

[…]

New York University neuroscientist Daniel Orringer and his colleagues developed a diagnostic technique that combined a powerful new optical imaging technique, called stimulated Raman histology (SRH), with an artificially intelligent deep neural network. SRH uses scattered laser light to illuminate features not normally seen in standard imaging techniques

[…]

To create the deep neural network, the scientists trained the system on 2.5 million images taken from 415 patients. By the end of the training, the AI could categorize tissue into any of 13 common forms of brain tumors, such as malignant glioma, lymphoma, metastatic tumors, diffuse astrocytoma, and meningioma.

A clinical trial involving 278 brain tumor and epilepsy patients and three different medical institutions was then set up to test the efficacy of the system. SRH images were evaluated by either human experts or the AI. Looking at the results, the AI correctly identified the tumor 94.6 percent of the time, while the human neuropathologists were accurate 93.9 percent of the time. Interestingly, the errors made by humans were different than the errors made by the AI. This is actually good news, because it suggests the nature of the AI’s mistakes can be accounted for and corrected in the future, resulting in an even more accurate system, according to the authors.

“SRH will revolutionize the field of neuropathology by improving decision-making during surgery and providing expert-level assessment in the hospitals where trained neuropathologists are not available,” said Matija Snuderl, a co-author of the study and an associate professor at NYU Grossman School of Medicine, in the press release.

Source: During Brain Surgery, This AI Can Diagnose a Tumor in 2 Minutes

The most direct and strongest evidence for the accelerating universe with dark energy is provided by the distance measurements using type Ia supernovae (SN Ia) for the galaxies at high redshift. This result is based on the assumption that the corrected luminosity of SN Ia through the empirical standardization would not evolve with redshift.

New observations and analysis made by a team of astronomers at Yonsei University (Seoul, South Korea), together with their collaborators at Lyon University and KASI, show, however, that this key assumption is most likely in error. The team has performed very high-quality (signal-to-noise ratio ~175) spectroscopic observations to cover most of the reported nearby early-type host galaxies of SN Ia, from which they obtained the most direct and reliable measurements of population ages for these host galaxies. They find a significant correlation between SN luminosity and stellar population age at a 99.5 percent confidence level. As such, this is the most direct and stringent test ever made for the luminosity evolution of SN Ia. Since SN progenitors in host galaxies are getting younger with redshift (look-back time), this result inevitably indicates a serious systematic bias with redshift in SN cosmology. Taken at face values, the luminosity evolution of SN is significant enough to question the very existence of dark energy. When the luminosity evolution of SN is properly taken into account, the team found that the evidence for the existence of dark energy simply goes away (see Figure 1).

Commenting on the result, Prof. Young-Wook Lee (Yonsei Univ., Seoul), who led the project said, “Quoting Carl Sagan, extraordinary claims require extraordinary evidence, but I am not sure we have such extraordinary evidence for dark energy. Our result illustrates that dark energy from SN cosmology, which led to the 2011 Nobel Prize in Physics, might be an artifact of a fragile and false assumption.”

Other cosmological probes, such as the cosmic microwave background (CMB) and baryonic acoustic oscillations (BAO), are also known to provide some indirect and “circumstantial” evidence for dark energy, but it was recently suggested that CMB from Planck mission no longer supports the concordance cosmological model which may require new physics (Di Valentino, Melchiorri, & Silk 2019). Some investigators have also shown that BAO and other low-redshift cosmological probes can be consistent with a non-accelerating universe without dark energy (see, for example, Tutusaus et al. 2017). In this respect, the present result showing the luminosity evolution mimicking dark energy in SN cosmology is crucial and very timely.

This result is reminiscent of the famous Tinsley-Sandage debate in the 1970s on luminosity evolution in observational cosmology, which led to the termination of the Sandage project originally designed to determine the fate of the universe.

This work based on the team’s 9-year effort at Las Campanas Observatory 2.5-m telescope and at MMT 6.5-m telescope was presented at the 235th meeting of the American Astronomical Society held in Honolulu on January 5th (2:50 PM in cosmology session, presentation No. 153.05). Their paper is also accepted for publication in the Astrophysical Journal and will be published in January 2020 issue.

Source: New evidence shows that the key assumption made in the discovery of dark energy is in error