Meniscal tear in the knee joint is a highly common injury that can require an ablation. However, the success rate of meniscectomy is highly impacted by difficulties in estimating the thin vascularization of the meniscus, which determines the healing capacities of the patient. Indeed, vascularization is estimated using arthroscopic cameras that lack of high sensitivity to blood flow. Here, we propose an ultrasound method for estimating the density of vascularization in the meniscus during surgery. This approach uses an arthroscopic probe driven by ultrafast sequences. To enhance the sensitivity of the method, we propose to use a chirp-coded excitation combined with a mismatched compression filter robust to the attenuation. This chirp approach was compared to a standard ultrafast emission and a Hadamard-coded emission using a flow phantom. The mismatched filter was also compared to a matched filter. Results show that, for a velocity of a few millimeters per second, the mismatched filter gives a 4.4-10.4-dB increase of the signal-to-noise ratio (SNR) compared to the Hadamard emission and a 3.1-6.6-dB increase compared to the matched filter. Such increases are obtained for a loss of axial resolution of 13% when comparing the point spread functions (PSFs) of the mismatched and matched filters. Hence, the mismatched filter allows increasing significantly the probe capacity to detect slow flows at the cost of a small loss in axial resolution. This preliminary study is the first step toward an ultrasensitive ultrasound arthroscopic probe able to assist the surgeon during meniscectomy.